Microscopic Properties of Asphalt and Polyethylene at an Extraordinary High Dosage through Molecular Dynamics Simulation

Abstract

Using waste plastics in asphalt mixtures could be an exploratory way to dispose of waste plastics. This study aims to investigate the microscopic properties between asphalt and polyethylene (PE) at an extraordinary dosage of 20 wt.%. Various types of PE with different degrees of polymerization (DP) and structural configurations were considered. Molecular dynamics simulations were used to calculate the mechanical parameters, free volume ratio (FVR), and Flory–Huggins parameter of the resulting PE-modified asphalt (PEA). Two types of PEA were made and characterized by fluorescence microscopy. The simulation results indicate that the addition of PE reduces the density of modified asphalt by less than 5%, and a higher density of PEA is associated with a lower FVR. When the FVR is close, the mechanical properties are greatly influenced by the DP and configuration. The DP and the number of chains are the main parameters impacting the compatibility between PE and asphalt, based on the Flory–Huggins parameter analysis. Decreasing the DP of PE (e.g., 50, with a minimum Flory–Huggins parameter and a relative molecular mass of 1300) will significantly increase the compatibility between asphalt and PE. LDPE−2 has better compatibility with asphalt, possibly because LDPE−2 has higher purity. These findings provide valuable insights into plastic thermal cracking and industrial modification practices.

1. Introduction

According to the Organization for Economic Co-operation and Development (OECD) research report released in 2022 [1], plastic production has increased to 460 Mt and plastic waste reached 353 Mt in 2019. Even during the COVID-19 pandemic and stagnation period, global plastic production increased to 390.7 million tons in 2021 [1]. However, only 15% of plastic waste was collected for recycling, and just 9% was recycled [2,3]. Millions of tons of recycled plastic waste are either landfilled or incinerated, causing secondary pollution to the environment and emitting significant amounts of carbon dioxide [3,4]. Microplastic particles also have a wide-ranging negative impact on human health, soil ecosystems, and aquatic ecosystems [5]. Plastic pollution has become the world’s second-largest environmental pollution problem after climate change.

Biodegradation is an environmentally friendly and promising method for plastic pollution control, such as using marine microalgae [6] and nematocyst venom protein [7]. The synergistic effect between microorganisms and their enzymes has been shown to enhance the efficiency of plastic degradation [8]. In the road industry, polymers have been widely used in the modification of asphalt since the second half of the twentieth century to decrease pavement deformation and premature rutting in heavy truck roads, especially in high-temperature areas [9]. Polyethylene (PE) is the most abundant and cost-effective recycled plastic, which can be melted and blended with asphalt materials at a standard mixing temperature. In essence, PE is expected to increase the rigidity of asphalt in high-temperature areas [10]. PE-modified asphalt (PEA) mixtures exhibit improved resistance to permanent deformation at high temperatures as well as thermal cracking under low temperatures [10,11,12,13,14,15].

PE is classified as a polyolefin in polymer and asphalt chemistry. Due to its nonpolar nature and high level of crystallinity, it can only be swelled by the maltenes of asphalt, resulting in the presence of a polymer-rich phase and an asphaltene-rich phase in PEA as observed through fluorescence microscopy [10,16,17,18,19,20]. The biphase nature of PEA blends always results in phase segregation due to differences in density, which poses a significant challenge to the storage stability of the binder during transportation to the application site. Early research has determined that incorporating 5% PE content is the optimal dosage to achieve outstanding performance of PEA [21,22,23,24,25], indicating that there are great difficulties in the large-scale application of PE in asphalt materials. The terminal effect has an important effect on the rheological properties of polymers [26]. To address this issue, researchers have employed chain group modification methods, such as NH₂-terminated PE [27], epoxy-terminated PE [28], or maleic anhydride grafted PE [29,30], to improve the storage stability of PEA. Additionally, the modification of polymer molecular architecture has been explored to promote the compatibility of PEA to some extent, including the molecular weight (MW) [31,32,33] and branch distribution [34,35] from experimental work. Despite numerous experimental efforts to modify PE, phase segregation and storage stability problems still limit the large-scale loading of PE in modified asphalt materials.

A key factor in improving the properties of PEA is understanding the microscopic interactions between polyolefins and asphalt. Previous studies suggested that intermolecular interactions within PEA primarily stem from PE and the aromatic phase of asphalt, as evidenced by direct fluorescent observation [9]. However, saturates, the least polar phase among the four components, also exhibit significant potential for miscibility with polyolefins. Classic atomic-level molecular simulation has become one of the commonly used methods to enhance our understanding of molecular processes in materials science [36]. With the aid of molecular dynamics simulation (MD) techniques, scholars can explore the microscopic mechanisms of polymer-modified asphalt at a molecular scale, following the construction of asphalt four-component models [37,38,39]. Recently, Hu et al. [40] demonstrated that graphene can effectively reduce the binding energy of asphaltenes to non-polar molecules (saturate and aromatic) and elevate the diffusion coefficient of non-polar molecules by approximately 8%. However, the diffusion coefficient of polar molecules is decreased by about 2% in modified asphalt systems with the addition of 5.8% PE. Yu et al. [41] utilized cohesive energy density (CED) as a crucial indicator to determine the degree of polymerization (DP) of PE molecules in natural rubber-modified asphalt. However, the best DP of PE found was only 12, which poses challenges for industrial applications.

In reality, PE or recycled PE typically has a high DP, with several thousand polymer chains splitting into an infinite number of short chains with different branches during utilization. Consequently, the use of PE or recycled PE is essential for modifying asphalt properties, and many researchers are currently exploring its large-scale collaborative utilization with other solid wastes, such as waste tire rubber [42]. Therefore, it is necessary to comprehend different PE configurations and DP using a large simulation system as a theoretical guide for PE thermal cracking or modification to enhance storage stability in modified asphalt.

To achieve large-scale use of waste plastic in road construction and consume more waste plastic products to reduce environmental pollution, an extraordinary high loading (20 wt.%) of PE was mixed with asphalt by using molecular dynamics simulation methods in this work. Different PE models were constructed with varying DP and branch arrangements for the first time to shed light on the influence of DP and molecular architecture on the performance development of PEA. The density, mechanical properties, free volume ratio (FVR), and mean squared displacement ($MSD$) of various types of PEAs were calculated. Their relationships were further analyzed. The compatibility between PE and asphalt was evaluated by calculating the Flory–Huggins parameter ($\chi$), which provided a theoretical basis for the chemical engineering modification of PE to improve its compatibility with asphalt in the future.

2. Computational Parameters and Experimental Methods

2.1. Molecule Models of Asphalt

The AAA-1 type asphalt in the Strategic Highway Research Program’s Materials Reference Library [43] was used in this study. Its molecular model was constructed with twelve component molecules based on Li and Greenfield’s study [37]. Materials Studio 2019 was employed for both model construction and MD simulation in this research. The molecular structures and component fractions of AAA-1 asphalt were described in our previous work [44]. The base asphalt in this work was composed of four component fractions, namely 16.6%, 10.7%, 30.7%, and 42.0%, which closely matched the experimental results [43].

2.2. Molecule Models of PE

PE is a high MW compound with a chemical formula of ${\left({\mathrm{C}}_2{\mathrm{H}}_4\right)}_{\mathrm{n}}$, ranging from several thousand to several million. In this work, 20 wt.% of PE was incorporated into the asphalt model. An AAA-1 asphalt conventional model has a relative MW of approximately 33,840.811, and the MW of the added PE was estimated to be around 8460. Therefore, according to this study, the added PE should have a DP of 300 in a conventional model.

The first parameter to consider is MW distribution, as it can affect the penetration, ductility, and softening point of modified asphalt [33]. Subsequently, different DP, number of branches (BN), and their arrangement in PE chains were taken into account during the model construction and labeled differently. Five chain types of PE were designed and labeled in

Table 1. Diagram for five chain types of PE.

Chain Type	Label	Chain Type
Single chain	SC
One branch	1B
Two branches	2B
Six branches distributed non-staggered	6BI
Six branches distributed staggered	6BII

, which are: single chain (SC), one branch (1B), two branches (2B), six branches distributed non-staggered (6BI), and six branches distributed staggered (6BII). Branched polymers can be visualized as linear polymers with side chains of the same polymer attached to the main chain. All the side chains were unified to have a DP of 10 and were distributed at 1/2, 1/3, or 1/4 of the main chain. Even though the DP and distribution of side chains in PE are more complex, this study can serve as a theoretical guide for modifying PE in real-world scenarios, as it explores the impact of different chain types on the compatibility between the PE chains and asphalt.

Afterward, a PEA model was created by adding 20 wt.% of PE compounds to predict its performance evolution. The structural properties of the PE compounds, including DP, CN, and BN, were taken into account shown in

Table 2. The combination of PE compounds with different DP, CN, and BN.

Combination	Degree of Polymerization (DP)	Number of Chains (CN)	Number of Branches (BN)	Branch Chain Ratio (%)
900-SC	900	1	0	0
600-SC	600	1	0	0
300-SC	300	1	0	0
300-1B			1	3.3
300-2B			2	6.6
300-6BI			6	20
300-6BII			6	20
150-SC	150	2	0	0
150-1B			1	6.6
150-2B			2	13.3
150-6BI			6	40
150-6BII			6	40
100-SC	100	3	0	0
100-1B			1	10
100-2B			2	20
100-6BI			6	60
50-SC	50	6	0	0
50-1B			1	20
50-2B			2	40

. A total of 17 types of PE compounds with a combined DP of 300 were designed in a single conventional cell. Additionally, a long PE chain with a DP of 600 or 900 was also constructed using two or three conventional cells to investigate the size effect and the case of a long PE chain.

2.3. Construction of PEA Models and Simulation Process

The research methodology is summarized in Figure 1. The simulation steps and parameter settings of PEA are described below:

(1)

Firstly, 19 types of PEA models were constructed using the COMPASS II force field through the ‘Amorphous cell module, and the initial density of models was set at 0.1 g/cm³. It should be noted that to create 600-SC or 900-SC modified asphalt, the conventional cell of asphalt was expanded to 2 or 3 to achieve a loading content of 20 wt.% of PE in the modified asphalt.

(2)

Then, the constructed PEA models were optimized by the Geometric optimization module with the following parameters: the energy convergence criterion was set as 1.0 × 10⁻⁴ kcal/mol, and the maximum force, maximum stress, and maximum displacement were set as 0.005 kcal/mol/Å, 0.005 GPa, and 0.001 Å, respectively.

(3)

Afterward, annealing was carried out under an NVT ensemble with an initial temperature of 298 K and a maximum temperature of 500 K, completing 10 cycles to release the energy generated during the modeling process for 1 ns.

(4)

In the dynamic simulation process, the 19 types of PEA models were simulated under an NPT ensemble for 500 ps at 298 K with a step of 1.0 fs to obtain a more stable structure. The pressure was set at 1.0 × 10⁻⁴ GPa. Finally, MD simulation in the NVT ensemble was performed at 298 K for 500 ps for each model, and all the analyzed data were based on these trajectories. The configuration of the 300-SC model is shown in Figure 2 after dynamic simulation.

2.4. Properties Calculation and Analysis

To further study the influence of DP and configuration of PE on the microscopic properties of modified asphalt, the mechanical properties, free volume ratio (FVR), Flory–Huggins parameter ($\chi$), and mean squared displacement (MSD) of the PEA models were calculated and the relationships among these parameters were discussed.

(1)

Mechanical properties

The Bulk modulus, shear modulus, and Young’s modulus of PEA were calculated to evaluate the resistance of PEA models to compressive deformation, shear strain, and tensile deformation. The Forcite module and constant strain method were used to calculate the mechanical properties of the models. The number of steps for each strain and the maximum strain amplitude were set as 4 and 0.003. The energy convergence criterion was set as 1.0 × 10⁻⁴ kcal/mol and the maximum force, maximum stress, and maximum displacement were set as 0.005 kcal/mol/Å, 0.005 GPa, and 0.001 Å. The Hill method was applied to compute the three parameters based on the elastic stiffness matrix and elastic compliance matrix of the modified asphalt models [45].

(2)

Free volume ratio (FVR)

Based on the Connolly surface method [46], the $FV$ and $OV$ were calculated using a Connolly probe with a radius of 1.00 Å. The Van der Waals scale factor was set to 1.00, and the grid interval was 0.75 Å. The $FV$ on the side of the atom volume surface is free of atoms, indicating the volume enclosed by the outside (blue-colored side) of the isosurface, as shown in Figure 3. The $FVR$ represents the ratio of the free volume to the total volume of the PEA models and can be calculated by Equation (1).

$$FVR=\frac{FV}{FV+OV}\ast 100\%$$

(1)

where $FVR, FV$, and $OV$ are free volume ratio, free volume (ų), and occupied volume (ų).

(3)

Flory–Huggins parameter (χ)

Two materials with similar solubility parameter values and low $\chi$ value may be miscible [47]. The Flory–Huggins parameter (χ) was used to evaluate the compatibility of PE with asphalt systems in this study. CED is the energy required by a unit volume of condensate to overcome the intermolecular forces, which is a measure used to evaluate the strength of intermolecular interaction [48]. The solubility parameter is the square root of the $CED$. CED and $\delta$ were calculated using Equations (2) and (3). The Flory–Huggins parameter (χ) was calculated by Equations (4) and (5) [49,50].

$$CED=\frac{E_{coh}}{V}$$

(2)

$$\delta =\sqrt{CED}$$

(3)

$$\mathsf{\Delta }{E}_{mix}={\phi }_A{\left(\frac{E_{\mathrm{coh}}}{V}\right)}_A+{\phi }_B{\left(\frac{E_{coh}}{V}\right)}_B-{\left(\frac{E_{coh}}{V}\right)}_{mix}$$

(4)

$$\chi =\left(\frac{\mathsf{\Delta }{E}_{mix}}{RT{\phi }_A{\phi }_B}\right)V_m=\frac{V_m}{RT}{\left({\delta }_A-{\delta }_B\right)}^2=\frac{M_n}{\rho RT}{\left({\delta }_A-{\delta }_B\right)}^2$$

(5)

where $E_{coh}$ is the cohesive energy (J), V is the volume of the system (m³), $\mathsf{\Delta }{E}_{mix}$ is the mixing energy of the mixture, ${\phi }_A$ and ${\phi }_B$ are the volume fractions of A and B in the mixed system, $V_m$ is the molar volume of asphalts, $R$ is the molar gas constant (8.314 J/mol/K), $T$ is the temperature of the system, $M_n$ is average molecular weight, $\rho$ is the density.

(4)

Mean squared displacement (MSD)

MSD is a parameter used to measure the deviation of the position of the particle with respect to the reference position after moving over time. The MSD (Ų) of each component in the PEA was calculated using the analysis function in the Forcite module, as shown in Equation (6).

$$MSD (t)=\frac{1}{N}<\sum _i^N{|r_i\left(t\right)-r_i\left(0\right)|}^2>$$

(6)

where $N$ is the number of diffused molecules in the system, $r_i\left(t\right)$ is the position of the particle at time $t$, $r_i$(0) is the position of the particle at time 0, < > is the average of all atoms in the group.

2.5. Experimental Part: Synthesis of PEA and Microscopic Tests

Two types of recycled low-density polyethylene (LDPE), namely LDPE−1 and LDPE−2, reproduced by sound-absorbing materials and packaging film, were used to synthesize PEA. These recycled plastics are produced by a qualified company and fulfilled with GB/4006-Plastics-Recycled plastic and SB/T 11149-2015-Technical specifications of waste plastics collection and sorting [51]. A Fourier transform infrared spectrometer, FTIR, (Nicolet 380, Attenuated Total internal Reflectance) and thermogravimetric-differential scanning calorimetry analysis, TG-DSC, (STA 449F5, 10 K/min under a nitrogen gas atmosphere) were used to characterize them. Base asphalt 70^# (Sinopec Zhenhai Refining & Chemical Company, Ningbo, China) was selected as the starting material to make LDPE-modified asphalt. The asphalt was heated to 150 °C and mixed with 20% LDPE using a high shear dispersing emulsifier (FM300) for 30 min, resulting in the production of PEA samples as shown in Figure 4. The microstructure of PEA was observed using a fluorescence microscope (OLYMPUS CKX53) with blue and green excitation light sources.

3. Results and Discussion

3.1. Density of PEA Models

After conducting a molecular dynamics simulation, the densities of 20 different asphalt models, including AAA-1, were compiled. As illustrated in Figure 5, the density of the AAA-1 asphalt model was the highest, and the densities of all PEA models were lower than that of AAA-1. The density of the LDPE modifier typically ranges from 0.91 to 0.94 g/cm³ [52,53], while the density of asphalt at room temperature is approximately 1.0 g/cm³ [44]. Therefore, the addition of 20 wt.% PE content will decrease the density of modified asphalt. Longer chain lengths of PE can further reduce the density of asphalt, with the density of PE at 900 DP and 600 DP being 0.91 and 0.95 g/cm³, respectively. The average density of PEA’s shorter chain lengths is similar to that of AAM-1, as calculated in our previous work using MD simulations [44]. This indicates that the composition has a significant impact on fundamental physical parameters. The densities of PEA samples with 300 DP, 150 DP, 100 DP, and 50 DP are 0.965 ± 0.004 g/cm³, 0.960 ± 0.004 g/cm³, 0.966 ± 0.006 g/cm³, and 0.959 ± 0.002 g/cm³, respectively. Generally, there is no significant correlation between densities and configurations in different types of PEAs with the same DP of PE. Among them, the density of 300-SC, 300-6BII, and 100-1B is relatively high.

3.2. Mechanical Properties of PEA Models

Table 3. Mechanical properties of pure PE, AAA-1, and PEA models with different chain types at 298 K.

Pure PE, AAA-1 and PEA Models with Different Chain Types	Bulk Modulus, B (GPa)	Shear Modulus, G (GPa)	Young’s Modulus, E (GPa)
900PE	2.844	0.715	1.979
300PE-3B	2.827	0.724	2.001
AAA-1	3.103	0.666	1.864
900-SC	2.509	0.570	1.589
600-SC	3.006	0.531	1.505
300-SC	3.165	0.779	2.159
300-1B	2.947	0.651	1.819
300-2B	3.110	0.731	2.035
300-6BI	2.992	0.726	2.014
300-6BII	3.172	0.833	2.297
150-SC	3.044	0.650	1.822
150-1B	2.947	0.736	2.038
150-2B	2.732	0.692	1.914
150-6BI	2.824	0.529	1.495
150-6BII	2.983	0.600	1.688
100-SC	3.048	0.545	1.542
100-1B	3.139	0.803	2.220
100-2B	3.062	0.680	1.898
100-6BI	3.014	0.649	1.817
50-SC	2.983	0.734	2.035
50-1B	3.087	0.755	2.095
50-2B	2.843	0.631	1.763

presents the mechanical properties of the 20 types of PEA models at a temperature of 298 K. Additionally, the mechanical properties of two pure PE models with a total DP of 900, namely 900PE and 300PE-3B, were calculated and compared with those of PEA. Generally, among the 19 types of PEA models, only the bulk moduli of 300-SC, 300-2B, 300-6BII, and 100-1B are higher than that of AAA-1. The shear moduli and Young’s moduli of 10 types of PEA models are higher than those of AAA-1, indicating that the addition of PE can improve the shear and Young’s modulus of PEA to some extent. Xia et al. [14] have previously demonstrated that PE with high crystallinity can enhance the stiffness of asphalt, thereby improving its rutting resistance through rheological temperature ramp tests and multiple stress creep recovery tests. This could be attributed to the high shear modulus and Young’s modulus of pure PE, which improves the mechanical properties of PEA, as shown in Table 3.

In general, four types of PEA, namely 300-SC, 300-2B, 300-6BII, and 100-1B exhibit better mechanical properties than AAA-1. Compared with AAA-1, the bulk modulus, shear modulus, and Young’s modulus of 300-6BII are increased by 2.22%, 25.08%, and 23.23%, respectively. Conversely, the bulk modulus of 900-SC is the worst, while the shear modulus and Young’s modulus of 150-6BI modified asphalt model are the worst. A low bulk modulus indicates that the model is less rigid and more prone to deformation, and 900-SC has the highest volume and lowest density (as shown in Figure 5) due to the incorporation of the longest chain length of PE with 900 DP.

Furthermore, data analysis from Table 3 reveals that the mechanical properties of 300-SC modified asphalt are superior to those of 600-SC and 900-SC modified asphalt when considering the same size ratio. This suggests that a smaller DP leads to greater stability in a system with the same MW ratio. In a single-chain combination, DP may be the primary parameter affecting the mechanical properties. When considering chain type analysis, except for the 300-DP combination, PEA with one PE branch chain exhibits the best shear modulus and Young’s modulus.

However, in general, the DP has a greater impact on the mechanical properties than the molecular architecture. Other factors, such as the length and location of branch chains, and the number of chains could also have some effects on the mechanical properties of asphalt with the same DP. The fundamental law governing these effects requires further study in the future.

3.3. Relationship between FVR and Mechanical Properties of PEA Models

According to Equation (1), the FVR for the 20 types of asphalt models were calculated and listed in

Table 4. FVR of AAA-1 and PEA models with different chain types.

AAA-1 and PEA Models with Different Chain Types	Occupied Volume (Å3)	Free Volume (Å3)	Free Volume Ratio (%)
AAA-1	45,096.22	8521.84	15.89
900-SC	175,207.85	48,425.93	21.65
600-SC	117,243.81	25,963.81	18.13
300-SC	58,714.22	11,571.54	16.46
300-1B	58,756.04	11,908.56	16.85
300-2B	58,888.77	11,868.92	16.77
300-6BI	58,792.02	12,181.74	17.16
300-6BII	58,847.59	11,463.19	16.30
150-SC	58,832.58	12,398.53	17.41
150-1B	58,630.93	12,060.54	17.06
150-2B	58,910.91	12,260.21	17.23
150-6BI	58,736.90	12,188.74	17.19
150-6BII	58,722.66	11,901.67	16.85
100-SC	58,713.87	12,131.01	17.12
100-1B	58,660.03	11,251.54	16.09
100-2B	58,899.58	11,715.55	16.59
100-6BI	58,702.49	12,054.54	17.04
50-SC	59,035.95	11,969.21	16.86
50-1B	58,638.13	12,540.28	17.62
50-2B	58,901.69	12,055.43	16.99

. Among the PEA models, 100-1B, 300-6BII, and 300-SC exhibit the smallest FVR. FVR has a low correlation with the DP and configuration of PE. Figure 6 illustrates a negative linear relationship (R² = 0.9784) between density and FVR, indicating that a higher density of PEA is associated with lower FVR.

Figure 7, Figure 8 and Figure 9 show the relationship between the FVR and bulk modulus, shear modulus, and Young’s modulus of different types of PEAs. Compared to other PEA models, 300-6BII, 300-SC, and 100-1B, which have the smallest FVR, exhibit better mechanical properties, suggesting that a smaller FVR leads to better resistance to compression deformation and shear strain. In the same size ratio of PEA combinations, 300-SC has a lower FVR and better mechanical properties than 600-SC and 900-SC. From the gray shaded area in Figure 7, Figure 8 and Figure 9, it can be seen that under similar FVR conditions, the bulk modulus, shear modulus, and Young’s modulus of different combinations are greatly affected by the DP and configuration of PE.

3.4. Compatibility Evaluation of PEA Models

The Flory–Huggins parameters were calculated to analyze the compatibility between PE and asphalt systems in different combinations. Figure 10 and Figure 11 show the Flory–Huggins parameters of PE and asphalt systems with different PD and the number of chains at 298 K. According to Figure 10, the smaller the DP, the relative MW of one PE molecule is closer to that of the four components of asphalt, and the compatibility between PE and asphalt is better. Meanwhile, Figure 11 shows that the more chains there are, the better the compatibility between PE chains and asphalt.

When a single PE chain with 300 DP is added, the compatibility between PE and asphalt in 300-6BII is the best, which corresponds to the result of mechanical properties. Although the $\chi$ value is large at 300 DP, the PE component in PEA plays a dominant role in resisting compression deformation and shear strain with a relatively small FVR. When two PE chains with 150 DP are added, the compatibility of PE chains with asphalt components is improved. The 150-SC has the best compatibility. After 100-SC, the compatibility of PE chains with asphalt components further improved, among which 100-2B and 50-SC have the best compatibility.

The above results show that the compatibility of different PEA combinations was not significantly affected by the configuration. The DP and number of chains are the main parameters impacting the compatibility between PE and asphalt.

3.5. Mean squared Displacement (MSD) Analysis of PEA Models

The MSD of PE in the PEA models is plotted in Figure 12. The MSD of 100-SC is the largest, while that of 150-6BI is the smallest. As a whole, the red lines are higher than the blue lines, indicating that the displacement of SC and 1B is greater than that of 6BI and 6BII. PE with fewer branch chains is more likely to move.

The MSD of five components of PEA models with polymerization degrees of 300 and 150 was calculated, as shown in Figure 13 and Figure 14. When the DP is 300, the MSD of PE in PEA is the highest, and saturates rank second. The saturates are a light component and are easily moved in the system. When the DP changes from 300 to 150, the MSD of PE is closer to that of saturates than others in asphalt, indicating that the displacement of PE and saturation fraction is close. At this time, the compatibility between the PE and the components in the asphalt system improves.

3.6. Experimental Part: Microscopical Analysis of LDPE and PEA

Figure 15 shows the FTIR spectrum of the two LDPE samples. The functional groups of LDPE-1 and LDPE-2 samples are similar and consistent with previous spectral results [6,54,55]. All the spectra showed the presence of main group peaks at 2915 cm⁻¹, 2848 cm⁻¹, 1472 cm⁻¹, and 717 cm⁻¹. It is worth noting that LDPE-1 has small protrusions at 1262 cm⁻¹, 1098 cm⁻¹, and 1026 cm⁻¹, suggesting the presence of C-O bonds in the sample. In LDPE-2, the vibration peak of the C-O bond is at 1020 cm⁻¹. The LDPE samples were prepared with recycled plastic particles, which suggested that they are oxidized at different levels.

The TG/DSC curves of LDPE-1 and LDPE-2 are shown in Figure 16. The melting point of LDPE is approximately 108–126 °C. Both LDPE-1 and LDPE-2 samples initially have an endothermic peak at 111.86 °C and 125.37 °C, which fall within the melting point range of LDPE. Comparing these two temperatures, the crystallinity of LDPE-2 is greater than that of LDPE-1. There is a significant exothermic peak at 450 °C of LDPE-2, indicating a secondary polycondensation reaction. The mass loss of LDPE-2 is higher at more than 500 °C, implying a higher purity of LDPE-2.

Two PEA samples amplified 10 times by fluorescence microscopy are shown in Figure 17. The π electrons in aromatic or double bond-containing compounds have conjugation, which can absorb light and enter the excited state, thus producing fluorescence [9,56]. In the mixed PEA sample, PE was excited after adsorbing aromatic or double bond-containing components in asphalt, resulting in a fluorescence effect. From Figure 17a,c, PE is dispersed with long chains stacked, while part component of asphalt is distributed on the PE surface. Phase boundaries between the polymer-rich phase and asphaltene-rich phase are clear. The Image J software 2021 [57] was used to analyze the mean fractal diameters of Figure 17b,d as 22.15 Å and 14.2 Å, respectively. A smaller mean fractal diameter indicates better compatibility, thus LDPE−2 has better compatibility with asphalt. The reason could be that LDPE−2 has higher purity.

Furthermore, different components in asphalt can be excited under different excitation light sources with various energies (blue or green light in this work). From Figure 17a–d, different fluorescence effects were observed in the same sample with blue or green light, showing that the PEA can be excited under green light more easily. Green light has a longer wavelength than blue light. It has been stated that the aromatic phase is more easily excited while the other components are more difficult. The characterization of asphalt with fluorescence microscopy is still progressing [56]. Figure 17b,d show the significant emission characteristics of the aromatic component and the PE mixed with the other components. Figure 17a,c further indicates that blue light with shorter wavelengths can excite the other components of asphalt, which brings a new method to distinguish different components of asphalt with fluorescence microscopy. But more work should be done in future. In summary, it is evident that PE and asphalt are still in a state of partial miscibility at the microscale from fluorescence microscopy observation.

4. Conclusions

This study aims to determine the microscopic properties of asphalt modified with PE at an extraordinary high dosage of 20 wt.% both from molecular dynamics and experimental works. Different DP and structural configurations (the number of branches and their arrangement in the main chains) of PE were designed and mixed with asphalt to build PEA molecular models. The density, mechanical properties, free volume ratio (FVR), mean squared displacement ($MSD$), and Flory–Huggins parameter ($\chi$) of various types of PEAs were calculated to analyze their relationships. Two types of PEA were made and characterized by fluorescence microscopy. Based on the results and analysis, it can be concluded that:

(1)

The addition of PE can reduce the density of modified asphalt by less than 5%, and a higher density of PEA is associated with lower FVR;

(2)

When the FVR is close, the mechanical properties are greatly influenced by the DP and configuration of PE.

(3)

The DP and the number of chains are the main parameters impacting the compatibility between PE and asphalt based on the Flory–Huggins parameter analysis. Decreasing the DP of PE (e.g., 50, with a minimum Flory–Huggins parameter and a relative molecular mass of 1300) significantly increases the compatibility between asphalt and PE.

(4)

LDPE-2 has better compatibility with asphalt, possibly because LDPE−2 has higher purity.

From this work, the mechanical properties of PE, the compatibility between asphalt and PE, and intermolecular interaction in the PEA blend system should have relationships, thus affecting the storage stability. Decreasing the DP of PE increases the compatibility between asphalt and PE, which could provide theoretical guidance for chemical engineering synthesis. Further research is needed to modify the PE chains to improve their blending ability with the four components of asphalt, thus increasing the storage stability of PEA. Their mechanical properties and binding ability with aggregates should also be investigated in the future. Furthermore, microplastic pollution generated by the widespread use of PEA in road engineering cannot be ignored.