3.1. Frequency Sweep Tests
The complex shear modulus master curves for A9, S4 and S4P3 before and after aging are shown in
Figure 1a. In general, the complex shear modulus decreased as the reduced frequency increased; besides this, the differences in all the binders gradually reduced when the frequency was around 10
3 to 10
9 Hz, and the shear modulus of all asphalt binders was finally around 10
9 Pa. According to the time-temperature superposition (TTS) principle, the higher the frequency response, the lower the temperature. This means that as the temperature decreased, the deformation resistance of the asphalt binders increased and the binders showed more elastic behavior. At lower frequencies (higher temperatures), A9 had the lowest complex shear modulus and S4 had the higher complex shear modulus. This indicates that, after adding the SBR, the high-temperature performance of the asphalt binder increased, and the complex shear modulus was further increased after adding PF; S4P3 exhibited the highest value of complex shear modulus. We also found that in the higher-frequency region (lower temperatures), the difference in the complex shear moduli of S4P3 and S4 was not significant. This indicated that S4P3 has a better high-temperature performance than A9 and S4, and there may be a slight effect of low temperature after adding PF. This will be further discussed in the temperature sweep. Overall, compared with the unaged asphalt binder, aging further increased the complex shear modulus, indicating that aging reduces flexibility. In addition, A9 had the highest complex shear modulus difference after aging, while the difference between S4 and S4P3 was lower. This shows that the aging resistance of the asphalt binder was enhanced after adding SBR and PF.
The phase angle master curves for A9, S4 and S4P3 before and after aging are shown in
Figure 1b. The phase angle can be used as an indicator of the viscoelastic behavior of asphalt binders; besides this, researchers have also found that the temperature of the phase angle at around 30° can effectively predict the low-temperature performance [
34].
Figure 1b shows that with increasing frequency, the phase angle gradually decreased, which means that as the temperature decreased, the elasticity of the asphalt binder gradually increased. Moreover, S4P3 showed the lowest phase angle value compared with S4 and A9 in a higher temperature region, and the difference in S4 and A9 was not obvious when the frequency was around 10
−1 to 10
2 Hz. A lower phase angle shows a higher elastic response; this therefore indicates that SBR resulted in a limited improvement on the high-temperature performance of asphalt binders, while PF could significantly improve the high-temperature performance of asphalt binders. Moreover, compared with other binders, S4P3 had a higher phase angle value in the high-frequency region (around 10
5 to 10
9 Hz), which means that S4P3 was more flexible under the same conditions.
3.2. Temperature Sweep Tests
To investigate the rheological properties of A9, S4 and S4P3 before and after aging under a certain temperature range (−20 to 80 °C), temperature sweep tests were performed. The test results are shown in
Figure 2a–d. As shown in
Figure 2a,b, S4P3 had the highest complex shear modulus and lowest phase angle value over the whole temperature range, which shows that S4P3 has a better high-temperature performance. Moreover, it was observed that there was no obvious difference in the complex shear modulus and phase angle between A9 and S4 when the sweep ranged from 60 °C to 80 °C. This indicates that in the high-temperature aspect, the improvement caused by SBR was limited.
For low-temperature sweep results are shown in
Figure 2c,d. Obtained results show that as the temperature decreased, the complex shear modulus increased while the phase angle decreased. As shown in
Figure 2c,d, S4 had the highest value of the complex shear modulus and lowest value of the phase angle when the temperature was lower than 0 °C. Moreover, when the sweep temperature was lower than −10 °C, the complex shear modulus and phase angle of the three asphalt binders tended to be stable. This means that the asphalt binders tended to be elastomers at lower temperatures, and the change in temperature had a slight effect on the modulus. Compared with S4, when the temperature was lower than −10 °C, the complex shear modulus of S4P3 was slightly lower than that of S4.
Generally, binders with higher complex modulus crack more quickly as these binders are unable to relax thermal stresses at low temperatures [
3]. Thus, S4P3 has better deformation ability at lower temperatures. In addition,
Figure 2d shows the slope of the phase angle of S4 and A9 changes steeper when the temperature exceeded −10 °C, while the change of S4P3 is slight. In general, the curvature of the phase angle can reflect the sensitivity of the material to temperature. Hence, it also means that S4P3 has a lower susceptibility to temperature. This is mainly because the embedded network structure formed by PF and SBR reduces the impact of temperature changes on the performance of asphalt and rubber particles. Furthermore, these embedded PF particles do not inhibit the deformation and recovery of rubber under the action of external stress, so S4P3 has better performance at lower temperatures.
3.3. Aging Index (AI)
To quantitatively analyze the effect of different aging degrees on the rheological properties of asphalt binders, AI was used to evaluate the aging degree, calculated by Equation (4). High values of this ratio indicate a relatively large degree of binder hardening [
35,
36]. The
G*/sin (
δ) was collected by the temperature sweep test with a controlled strain at 10%, and the loading frequency was 10 rad/s.
where
G*/sin (
δ)
Aged is the asphalt binder rutting parameter values obtained for the aged material and
G*/sin (
δ)
Unaged is the asphalt binder rutting parameter values obtained for the unaged material.
The results are shown in
Figure 3 and
Table 3 and show that the AI increased as the aging time increased for all the binders. Furthermore, as the aging time increased, the degree of aging increased in a linear fashion. Moreover, A9 had the highest AI value and slope, and S4P3 had the lowest value. When the aging time was changed from one hour to three hours, the AI index of S4 and S4P3 increased by 157% and 86%, respectively. This indicated that S4P3 has better aging resistance than S4, and the aging time has less influence on its rheological properties. This is mainly because a stable cross-linking system was formed by PF and SBR, which inhibited the volatilization of light components, and thus the aging resistance of S4P3 was improved. However, it should be noted that the difference between S4 and S4P3 was not significant for a short period of aging. As the aging time increased, the difference in the AI between the S4P3 and S4 also gradually increased, and this difference was obvious when the aging time was 7 h. This indicated that, in the initial aging stage, due to the swelling and adsorption of rubber particles in the asphalt, the aging resistance of the asphalt binder was improved to a certain extent, but with the increase in the aging time and aging degree deepening, the rubber particles also aged and disintegrated. Thus, the aging resistance of asphalt gradually weakened.
3.4. FTIR Analysis
The characteristic peak assignment of the spectra for the binders is shown in
Table 4. To investigate the PF modification mechanism, a FTIR test was also conducted on pure PF-modified asphalt (3PF). The FTIR spectra of A9, S4 and S4P3 before and after aging are given in
Figure 4,
Figure 5,
Figure 6 and
Figure 7, respectively.
The FTIR spectra of A9, S4, 3PF and S4P3 are shown in
Figure 4. At 967 cm
−1, S4 exhibited a new absorption peak due to the SBR characteristic peak vibration. Moreover, no new peak was generated compared with the A9 spectrum. This indicated that the modification mechanism of SBR was mainly physical modification without a chemical reaction. Compared with 3PF and S4, new absorption peaks appeared at 967 cm
−1 and 695 cm
−1, and the intensity of the absorption peak at 1515 cm
−1 increased significantly. This indicated that the functional groups in the asphalt were transformed after adding PF, and more stable C-C bonds were generated. This also means that the modification mechanism of PSBR is a physical modification and is accompanied by chemical reactions.
The FTIR spectrum of A9 at different aging times is shown in
Figure 5. During the whole aging process, only the intensity of the absorption peaks changed significantly. This indicated there were no functional group changes and generations, but some component transfers were present. Especially for the absorption peaks at 1604 cm
−1 and 1024 cm
−1, their intensity changed significantly after aging for one hour. This showed that the chemical system of A9 is more sensitive to aging, and more carbonyl groups and sulfoxides will be formed in the early stage of aging.
Figure 6 shows the chemical bond changes in S4 at different aging times. After one hour of aging, a new absorption peak appeared at 849 cm
−1, and the intensity of the absorption peak also increased as the aging time increased. Moreover, when the aging time was 7 h the intensity increase was most obvious and a new absorption peak appeared at 940 cm
−1. This indicated that S4 was more sensitive to aging, and chemical reactions occurred in the initial stage of aging, so some chemical bonds were broken and new chemical bonds were formed. This is mainly because a higher number of butadiene structures contained in the SBR molecule makes it easier for it to be oxidized or decomposed in short-term aging.
The FTIR spectrum of S4P3 is shown in
Figure 7. Before 5 h of aging, there were no significant changes in the absorption peaks; the strength changed as the aging time increased. This indicated that, before 5 h of aging, S4P3 could keep the chemical system stable with better thermal aging resistance than S4. When the aging time was 5 h, the absorption peaks changed clearly at 940 cm
−1 and 849 cm
−1. In addition, with the aging time increasing to 7 h, the intensity of the absorption peak at 940 cm
−1 and 1520 cm
−1 increased significantly, and the absorption peak at 848 cm
−1 was broadens.
It indicated that when the aging time exceeds 5 h, the chemical system of S4P3 became non-stable, and these peaks transferred to another functional group. Thus, S4P3 has good thermal aging resistance within 5 h of aging; after that, with the aging time increasing, the thermal aging resistance of S4P3 also degrades, which lead to make the chemical bond became non-stable.
To quantitatively analyze the functional group evolution of asphalt binders at different aging degrees, the structural indices
IC=O,
IS=O,
IC=C, and polymer index (
IBI) were used [
37,
38]. The
IC=O,
IS=O,
IC=C, and
IBI were calculated by using Equations (5)–(8), respectively, as presented in
Figure 8a–d.
Figure 8a,b show the carbonyl and sulfoxide evolution at the different aging times, respectively. In general, as the aging time increased, the carbonyl and sulfoxide rapidly increased, except the carbonyl of S4P3 at the initial stage decreased from unaged to 1 h of aging. Moreover, it was found that A9 had the highest rate of formation of carbonyl compounds, which was less in S4 and S4P3 when compared with A9; besides this, S4P3 had the lowest rate compared with other binders. This means that aging has a significant effect on A9, and as the aging time increases, the chemical reaction gradually increases. Moreover, the thermal aging resistance of A9 increased after being composited with SBR, but this improvement was limited and S4P3 had better thermal aging resistance compared with S4. It is well known that the better dispersibility of asphaltenes leads to increased formation of carbonyl compounds [
39]. However, after the addition of PF and SBR, the asphaltene network of the base asphalt was disturbed; thus, carbonyl formation was reduced. A clear increase in sulfoxide was observed for all binders; S4 had a higher slope compared with A9 and S4P3. These increases could be due to the reaction of oxygen with the perhydroaromatic ring to form hydroperoxides, and thus ketones and sulfoxides were formed. Moreover, after 7 h of aging, the sulfoxide compounds of A9 were reduced. This is attributed to the decomposition of sulfoxides after prolonged periods of aging [
40,
41].
Figure 8c shows the aromatic index evolution for different aging times. We can observe that the evolution of the aromatic index for all the asphalt binders was greatly different, especially for S4 compared with A9 and S4P3. In the initial aging stage, the aromatic fraction increased for both asphalt binders. This increase is certainly the result of the aromatization of the naphthenic compounds [
42]. After one hour of aging, the increase rate of the aromatic fraction of S4 was decreased; on the contrary, the aromatic compounds of A9 and S4P3 were decreased. The increase rate of aromatics of A9 and S4P3 rapidly increased from 3 h to 5 h. The same phenomenon was also found by Nivitha [
39]. Therefore, during the aging process, the addition of modifiers will cause different changes in the composition of asphalt, and this change will vary greatly due to the different modifiers, especially for the aromatic fraction. In summary, it is reasonable and accurate to use the carbonyl index to evaluate the aging effect of asphalt.
Figure 8d shows that, with increasing aging time, the
IBI of S4 and S4P3 both decrease. The
IBI of S4 changes significantly in the early stage of aging, while the change in S4P3 is relatively small. The
IBI in asphalt can reflect the change in the rubber particles. This means that rubber also deteriorates during the aging process, and the addition of PF reduces the degradation of the rubber during the aging process.
3.5. GPC Analysis
The molecular weight and molecular weight distribution of the asphalt are the inherent reasons for the performance of the asphalt, and thus have a great impact on its performance. Therefore, in this study, GPC was conducted to investigate the evolution of the molecular weight of the asphalt binders during the aging process to analyze the aging effect [
43,
44].
Figure 9 and
Table 5 shows the GPC results of asphalt binders before aging. We found that 3PF is mainly composed of small molecules (SMS) and medium molecules (MMS), as we expected. Compared with S4, the contents of large molecules (LMS) and SMS of S4P3 were increased. This is due to the entanglement and cross-linking between PF and rubber, which increased the overall average relative molecular mass. Our research found that the content of LMS in asphalt has a good correlation with the high and low-temperature performance of asphalt [
45]. Therefore, PF improves the high-temperature performance while maintaining better low-temperature performance, which also verified the results of the temperature sweep test results.
Figure 10a and
Table 6 show the molecular weight distribution evolution of A9 at different aging times. In general, as the aging time increased, the MMS were converted to LMS and SMS during the aging process, and the content of LMS and SMS increased gradually. It is worth noting that, when the aging time was 7 h, the LMS decreased compared with 5 h of aging. This may be due to excessive aging, which caused the LMS to decompose.
Figure 10b and
Table 6 shows the molecular weight distribution evolution of S4 at different aging times. As the aging time increased, the SMS and MMS were gradually converted to larger molecules. After one hour of aging, the content of LMS in the S4 increased significantly. This was due to the continued swelling of the rubber particles increasing the cross-link density and increasing the average molecular weight. Compared with 1 h of aging, the LMS in S4 was decreased after 3 h aging. This may be due to the rubber particle degradation into SMS under the influence of aging. It also means that S4 has poor aging resistance. Even when subjected to short-term aging, significant oxidation reactions occurred inside. When the aging time exceeded 5 h, the LMS content of the asphalt increased again. This was due to the deepening of the aging degree and the polymerization of more molecules in the asphalt.
The molecular weight distribution evolution of S4P3 at different aging times is shown in
Figure 10c and
Table 6. S4P3 shows significantly different changes in the molecular weight distribution evolution compared with S4. In general, during the whole aging process, the changes in the MMS in the S4P3 asphalt are more obvious, and the MMS are converted into LMS and SMS. Thus, the content of LMS and SMS in S4P3 was increased. The increase in LMS and SMS also contributed to the improvement of the performance of S4P3. Moreover, when the aging time was less than 5 h, there was no obvious degradation of the rubber particles, which also shows that the addition of PF can improve the thermal aging stability of asphalt binders.
3.6. Morphology Evolution
Asphalt and polymers reflect different colors under different light sources under a fluorescence microscope. Therefore, the dispersion state and particle size of the polymer in the asphalt can be observed. The morphological structures of S4 and S4P3 under different aging degrees are shown in
Figure 11a–j. The brighter part is the color reflected by the polymer, and the asphalt is the darker part of the figure.
From
Figure 11a,b, the size and dispersion of rubber particles in S4 and S4P3 can be seen to be roughly the same, and they have a better interfacial phase with asphalt. This indicates that the addition of PF will not affect the dispersion of SBR in asphalt and has good compatibility with asphalt. In S4P3, PF was dispersed between the asphalt’s continuous phase and rubber phase in a smaller size and formed an embedded network structure. PF showed better high-temperature performance and dimensional stability; thus, it played a key role in protection and thermal isolation, reducing the effects of temperature changes and aging on the asphalt and rubber particles.
The morphological evolution of S4 and S4P3 under different aging times is shown in
Figure 11c–j. It can be seen that the size and dispersion states of the polymer in S4 and S4P3 were significantly changed as the aging time increased. Aging had the most obvious impact on S4. As the aging time increased, the size of rubber particles in S4 decreased significantly. When the aging time exceeded 3 h, the degradation of rubber particles in asphalt was the most serious. When the aging time was 7 h, the rubber phase almost completely disappeared in the asphalt. This also shows that aging has a greater impact on rubber particles, as even a short aging time will cause significant changes, which again verifies the conclusions obtained in FTIR and GPC.
Compared with S4, S4P3 showed better stability during the whole aging process, and its morphological characteristics were slightly changed. In the early stage of aging (aging for one hour), the size of rubber particles decreased slightly, and the interface strength between PF and SBR increased. As the aging time increased, the size of the rubber particles in the internal system gradually decreased, but the rate of change was relatively slow. The rubber particles still maintained a good core after aging for 3 h, PF and SBR were well dispersed in the asphalt at the transition interface and the intensity increased. When the aging time exceeded 3 h, the PF began to solidify and polymerize, changing from a well-dispersed small particle state to a larger size particle, and the size of the rubber particles also began to decrease significantly. When the aging time was 7 h, PF continued to solidify and polymerize, and obvious continuous phases were formed between the PF itself and rubber particles. Therefore, the restriction of the asphalt molecules as further strengthened, and the asphalt exhibited better resistance to deformation; however, the relative deformation ability decreased significantly.