*3.4. Storage Stability*

Modified asphalt binder requires excellent storage stability during high-temperature storage, pumping, transportation and construction [55]. However, phase separation often occurs due to the incompatibility between polymers and asphalt binder. The principal reason for this is that the neat binder and polymer have great differences in density, molecular weight, polarity and solubility parameters. The phase separation leads to performance degradation in terms of anti-rutting, anti-cracking and fatigue resistance of modified binder, hindering the practical application of modified asphalt binder [56,57]. Therefore, the storage stability of composite modified binders containing various contents of APAO and PPA is evaluated by the storage test.

The temperature differences between the top and bottom parts in the storage test of modified binders are calculated, as shown in Table 9. To ensure the storage stability of the modified binder at a high temperature, a temperature difference of less than 2.5 ◦C is required. It is observed that the modified binder A0P1.5 is stable at elevated temperatures. After adding APAO, the temperature difference increases with increasing concentration of APAO, and phase separation becomes predominant for contents higher than 4 wt.%. Modified binder A2P0 has good storage stability due to the uniform dispersion of APAO in asphalt binder. After PPA is added, the softening point difference is less than 2.5 ◦C when PPA content reaches 2.0 wt.%. Such a phenomenon reveals that PPA has little influence on the storage stability of modified binders. In conclusion, the APAO proportion should be maintained at 2 wt.% to ensure storage stability.

**Table 9.** Softening point differences of composite modified binders.


It is generally believed that good compatibility between binder and polymer is essential to ensure the excellent stability of polymer modified binder. In view of the accuracy, the rheological method has a unique advantage in determining compatibility because it is sensitive to the differences between polymers and asphalt binder. At present, rheological methods such as master curves, black curves, Ham diagrams, and Cole–Cole diagrams are employed to characterize the compatibility of modified binders [54,55,58,59]. Among the above approaches, the Cole–Cole diagrams have proven to be the most effective method for determining compatibility. Thus, the compatibilities of APAO/PPA modified asphalt binder with various amounts of APAO and PPA were studied by a Cole–Cole diagram. The Cole–Cole plots comprise two variables, η and η", which are separated from the complex viscosity η\*. The compatibility is determined by the shape of the Cole–Cole diagram. A more symmetrical parabola represents better compatibility, while an asymmetrical parabola indicates poor compatibility.

The Cole–Cole diagrams of APAO/PPA modified binders at 60 ◦C are displayed in Figure 15. It is clear that the η" values of samples initially increase and then decrease with increases in η , with a peak value appearing on the curve. The modified asphalt binders with various concentrations of APAO and PPA exhibit different curve shapes, which mainly depend on the effects of the modifier on the viscoelastic constituents of the asphalt binder. The left part and right part of the Cole–Cole diagram characterize the elastic and viscous performances of the binder, respectively. As shown in Figure 15a, the modified binder A0P1.5 exhibits a symmetrical shape, indicating better compatibility between PPA and neat binder. After APAO is added, the curves gradually transition from right to left with increasing content of APAO, showing that APAO can enhance the elastic components of modified asphalt binder. The curve is mostly concentrated on the left side when the APAO content reaches 6 wt.%, suggesting that the elastic component occupies a dominant position. Moreover, the increase in PPA content also leads to improvement in elastic constituents, as demonstrated by the transition of curves from right to left. Remarkably, modified binder A2P1.5 shows the most symmetrical shape, followed by the modified binder A4P1.5. When the APAO concentration reaches 6 wt.% or the PPA content reaches 2.0 wt.%, the Cole–Cole curves show a tendency to deviate from the semicircle. This demonstrates that 2 wt.% APAO and 1.5 wt.% PPA are the optimal concentrations for the preparation of modified asphalt binder with superior compatibility.

**Figure 15.** Cole–Cole diagrams of composite modified binders: (**a**) composite modified binders with various contents of APAO; (**b**) composite modified binders with various contents of PPA.

It is important to investigate the distribution of polymer in asphalt binder to explore the behavior of polymer modified binders. The morphologies of APAO/PPA modified binders were observed by a fluorescence microscope. Figure 16 presents the fluorescence images of tested samples where the APAO particles are bright yellow and the asphalt binder phase is black. As observed in Figure 16a, the APAO particles appear in the form of spherical particles and are sparsely dispersed in the binder. However, the dispersion state of APAO changed significantly after the addition of PPA. The particle size decreases markedly and the distribution becomes much more uniform with increasing content of PPA, as shown in Figure 16b–d. This indicates that PPA is conducive to dispersing the APAO particles, and proves that PPA strengthens the compatibility between APAO and binder. Furthermore, Figure 16c,e,f show that the particle size of APAO increases and the distribution tends to aggregate as the APAO concentration increases from 2 wt.% to 6 wt.%. When the APAO proportion exceeds 4 wt.%, the boundary between the APAO and the asphalt binder phase is obvious and the dispersion worsens, revealing that potential phase separation may occur. This phenomenon corroborates the results of the Cole–Cole plots.

#### *3.5. FTIR Spectra*

The analysis conducted above shows the variations of the conventional and rheological performances of modified binder after modification with APAO/PPA. More importantly, the microstructure of the modified binder primarily affects its macroscopic rheological behavior. FTIR has been widely used as an efficient method to evaluate the variations in chemical bonds and structures of modified binders. To ascertain the modification mechanism of the APAO/PPA modified binder, the infrared spectra of neat binder and APAO/PPA modified binder were acquired. Figure 17 displays the FTIR spectra of the modified binders.

**Figure 16.** Morphology of tested samples (×400). (**a**) A2P0; (**b**) A2P1.0; (**c**) A2P1.5; (**d**) A2P2.0; (**e**) A4P1.5; and (**f**) A6P1.5.

**Figure 17.** Infrared spectra of tested samples.

It is observed that the typical characteristic peaks at 2956 cm−1, 2923 cm−<sup>1</sup> and 2853 cm−<sup>1</sup> denote the C-H stretching vibrations of aliphatic hydrogen. The absorption bands appearing at 1600 cm−<sup>1</sup> correspond to the stretching vibrations of C=C. The C-H bending vibrations are located at 1460 cm−<sup>1</sup> and 1376 cm<sup>−</sup>1. The characteristic absorption peaks within the 700–900 cm−<sup>1</sup> range are the C-H bending vibrations of the aromatic ring.

The APAO is formed through the polymerization reaction of α-olefines, and contains a polymerized methylene group. The 1160 cm−<sup>1</sup> and 972 cm−<sup>1</sup> peaks represent the wagging and rocking vibration of C-H, respectively. The methylene group corresponds to the absorption band at 722 cm<sup>−</sup>1. The above peaks are considered to be typical characteristic peaks of APAO [7]. It is clear that there is no difference in absorption peaks between the neat binder and modified binder A2P0 in the infrared spectrum, indicating that APAO reacts with asphalt binder physically. However, after PPA is added, a new absorption band appears at 1012 cm<sup>−</sup>1, representing the P-O stretching vibration [17,60]. Masson et al. [61–64] showed that a phosphorylated product was generated by the reaction of asphalt binder with PPA. Ge et al. [65] pointed out that PPA caused the generation and disappearance of some functional groups. These results reveal that PPA reacts with asphalt binder chemically. In conclusion, APAO and PPA modification involves both physical and chemical interactions.
