Abstract This paper aims to analyse the mechanisms of microscopic diffusion and adhesion at the interface between polymer-modified asphalt and aggregate at high blending ratios. Using molecular dynamics simulations, a model of the polymer-modified asphalt–aggregate interface was developed. The study systematically investigated the effects of three types of modifiers—SBS, SBR and PE—on the wetting, diffusion and adhesion behaviour at the asphalt interface within a content range of 2.5% to 10% and under various temperature conditions. The results indicate that an increase in modifier content inhibits the migration of light fractions towards the interface, leading to weakened interfacial diffusion and non-uniform wetting under high-temperature conditions. In contrast, the SBS and SBR systems are more sensitive to changes in temperature and modifier content, while the PE system exhibits a relatively weaker diffusion attenuation effect. The interfacial adhesion results further indicate that the adhesion energy between modified asphalt and aggregate is higher at low temperatures and is primarily governed by van der Waals forces; as the modifier content increases, interfacial adhesion at low temperatures is enhanced, but construction and coating performance at high temperatures are somewhat affected. This study elucidates the microstructural diffusion and adhesion mechanisms of high-content polymer-modified asphalt under cold-climate conditions, providing a theoretical basis for the selection of modified asphalt materials and the optimisation of mix designs in low-temperature regions. 1. Introduction In cold regions, asphalt pavement projects are often more prone to defects such as cracking and potholes under extremely low temperatures than they are other regions. This is essentially because asphalt, being a highly temperature-sensitive material, exists in a glassy state at low temperatures, a ductile-brittle state at ambient temperatures, and a viscous-flow state at high temperatures. When temperatures are low, it remains in the glassy state for extended periods, a state characterised by hard and brittle mechanical properties [ 1]. This reduces its ability to deform and adapt, as well as its adhesion to aggregates, making the asphalt mixture more prone to cracking. Based on this, this paper proposes the use of molecular dynamics (MD) simulations to analyse the mechanisms of microscopic diffusion and adhesion at the interface between polymer-modified asphalt and aggregate under high polymer content. A molecular interaction model between modified asphalt and aggregate was established, and the interaction effects between modified asphalt and aggregate under low- and high-temperature conditions were characterised using indicators such as diffusion coefficients, adhesion energy, and diffusion concentration distributions. This paper compares the performance of three modifiers—SBS, SBR and PE—at mixing temperatures and under low-temperature conditions across a wide range of dosage levels, thereby providing a more targeted theoretical basis for the design of asphalt materials in cold regions [ 10]. 2. Materials and Methods 2.1. Materials and Molecular Models The molecular structures of the three polymer modifiers are shown in Figure 2. For the molecular simulations, representative fragments were selected: for SBS and SBR, a block ratio of 3:7 (i.e., m/n = 3/7) was adopted, while for PE, a fragment of length 7 was used. The aggregate surface was constructed as a hydroxylated mineral slab to better represent the chemically active surface of real aggregates after environmental exposure. Surface hydroxyl groups were introduced on the outer atomic layer before interface construction, and the hydroxylated surface was subsequently energy-minimised to remove unrealistic dangling bonds and high-energy artefacts. The basic structure of this is shown in Figure 3. Real aggregate surfaces are rarely chemically inert. Surface hydroxylation introduces polar active sites that can participate in hydrogen bonding, electrostatic interactions, and specific adsorption with asphalt molecules. Therefore, a hydroxylated surface is more representative of practical aggregate interfaces than an ideal pristine mineral surface. 2.2. Methodology 2.2.1. Basic Approach to MD This paper employs MD simulations to investigate the microscopic adhesion mechanisms at the modified asphalt–aggregate interface. The basic procedure comprises five steps: molecular model development, model optimisation, model validation, MD simulation and data analysis, as shown in Figure 4. The development of a molecular model forms the foundation of all MD simulation studies. The purpose of model optimisation and validation is to ensure the accuracy of the molecular model; only a microscopic model that accurately reflects the properties of asphalt can guarantee the reliability of the computational results. Once the reliability of the model has been confirmed, various loads and boundary conditions (such as temperature and pressure) are applied to perform a kinetic analysis of the molecular simulation [ 15]. This ultimately yields the evolution patterns of various parameters during the MD simulation process, thereby reflecting the interaction behaviour between modified asphalt and aggregate. 2.2.2. Construction of MD Models To investigate the microscopic adhesion properties between modified asphalt and aggregates, this study first constructed molecular models of the modified asphalt and mineral aggregates using the Amorphous Cell and Surfaces modules in the MD simulation software Materials Studio 2019. The Layer module was then used to assemble the modified asphalt and mineral aggregates into an interlayered model, as shown in Figure 5. The specific number of asphalt molecules is shown in Table 4. During the modelling process, the general-purpose polymer force field Compass II was employed, and the model’s density was set to 1.0 g/cm 3. The upper layer of the layered model consists of modified asphalt, while the lower layer comprises mineral aggregate. As the model features a periodic structure, a 20 Å vacuum layer was introduced in the upper layer to isolate it and prevent interference from external molecules. The final dimensions of the layered model were 35 × 35 × 146 Å. SBS, SBR and PE were studied as three separate systems, corresponding respectively to SBS-modified asphalt/aggregate, SBR-modified asphalt/aggregate and PE-modified asphalt/aggregate; they were not added simultaneously to the same system. The modifier dosage was defined as the mass fraction of modifier in the total modified asphalt system. A single molecular chain does not correspond to a fixed dosage percentage. The target mass fraction was achieved by adjusting the number of modifier chains in the simulation cell according to the molecular weight of each modifier. In molecular dynamics simulations, the target concentration is approximated using an integer number of molecular chains. As the number of molecules in an MD model must be an integer, the number of molecules shown in Table 5 corresponds to the number of chains closest to an integer, calculated on the basis of molecular mass. Once the MD model has been established, geometric optimisation and annealing are required to ensure the rationality of the internal molecular structure and the stability of the system’s internal energy [ 16]. The model is optimised using the Smart algorithm, followed by 25 rounds of annealing. The system enters a 100 ps pre-equilibration phase, during which the total energy, temperature and density gradually stabilise. This indicates that the model has eliminated the non-physical voids and local stresses arising from the initial configuration, thereby reaching a state of equilibrium suitable for subsequent analysis. The subsequent 100 ps was designated as the production phase, during which data relating to diffusion and adhesion were collected under the same Isothermal–isobaric ensemble (NPT) conditions. 2.2.3. Validation of the MD Model To validate the validity of the asphalt molecular model, the density, solubility parameters and energy change parameters during the optimisation process were extracted from the model. Data were extracted during the production phase of the NPT system simulation to calculate solubility parameters. Density reflects the discrepancy between the model’s density and the measured value; the densities of the asphalt molecular model and the measured sample were 0.985 g/cm 3 and 1.011 g/cm 3 respectively, with a relatively small error between the two. The solubility parameters, meanwhile, reflect the compatibility between the various components and demonstrate the stability of the mixed model. The solubility parameters for the various asphalt components, as calculated, are shown in Table 5. It can be observed that the maximum difference in solubility parameters among the asphalt components is 0.833, which does not exceed 4; this therefore demonstrates that the molecules of the various asphalt components are highly compatible [ 17, 18]. The density and solubility parameters confirm that the constructed model accurately reflects the physical properties of asphalt. The solubility parameter is calculated using the following formula: C E D = H v − R T V , (1) δ = C E D , (2) Since the four components of asphalt naturally coexist in actual asphalt colloidal systems, the discussion in this paper does not concern the issue of compatibility—namely, whether the components can coexist—but rather whether the parameter distributions of the representative molecules in the constructed model are reasonable, and whether the model as a whole can adequately characterise the thermodynamic properties of actual asphalt systems. The results indicate that the solubility parameters of the various components differ only slightly, suggesting that the constructed model exhibits good thermodynamic consistency and can be used for subsequent analyses of interfacial diffusion and adhesion behaviour. The four-component, twelve-molecule model employed in this study is a representative simplification and cannot fully capture the entire chemical diversity of real asphalt. As the composition of real asphalt is influenced by the source of the crude oil, the processing methods and the degree of ageing, the model results are suitable for trend analysis and mechanistic interpretation rather than for the precise prediction of absolute values. The model was comprehensively evaluated in terms of density, solubility parameters and energy convergence during the geometric optimisation process to validate the validity of the constructed asphalt molecular model. The model density is relatively close to the measured asphalt density, indicating that the model provides a good representation of macroscopic volume characteristics; the solubility parameters of the representative molecules show little variation, suggesting good consistency in the thermodynamic properties of the model’s internal components; during the geometric optimisation process, the system’s energy decreased rapidly and stabilised, indicating that non-physical stresses and voids in the initial configuration had been effectively eliminated. These three indicators corroborate one another, demonstrating that the established asphalt molecular model can effectively reflect the fundamental structural characteristics and thermodynamic behaviour of actual asphalt systems, and can be utilised for subsequent analyses of interfacial diffusion and adhesion properties. 2.2.4. MD Simulation During the MD simulation, the number of particles, pressure and temperature of the system were kept constant; therefore, the ensemble was set to NPT. During the MD simulation, the number of particles, pressure, and temperature of the system were kept constant; therefore, the NPT ensemble was adopted throughout the simulation. This choice was used to relax the initial internal stress of the layered interface and to allow the density of the modified asphalt–aggregate system to evolve naturally toward a physically realistic state. To reduce the influence of transient fluctuations, only the equilibrated trajectory window was used for subsequent statistical analyses. With regard to the interfacial diffusion, wetting and adhesion behaviour investigated in this study, NPT conditions better reflect the volumetric response of materials under external temperature and pressure during actual paving and service, thereby enhancing the physical plausibility of the simulation results and the comparability between different operating conditions. The force field used was Compass II, and the temperature control method was the Berendsen method. The pressure was set to standard atmospheric pressure. The simulation time step was 1.0 fs, and the total number of time steps for the 200 ps simulation was 200,000. This paper primarily analyses the interaction mechanisms between high-content modified asphalt and mineral aggregate under low-temperature conditions and mixing temperatures. Consequently, −40 °C, −20 °C and 0 °C were selected as the simulation temperatures for the low-temperature environment, whilst 160 °C and 180 °C were chosen as the simulation temperatures for the mixing environment. The modifier content levels were set at 2.5%, 5%, 7.5% and 10%, respectively [ 22]. The purpose of MD is to observe residual molecular motion and interfacial rearrangement that persist at the nanoscale; it does not require asphalt to exhibit fluidity at the macroscopic level. Temperatures of −20 °C and −40 °C represent typical low-temperature and extreme low-temperature conditions in cold-region service environments, respectively. These allow for the observation of gradient patterns in diffusion and adhesion changes during the cooling process; even at −40 °C, limited molecular migration can still be observed within the MD timescale, which is meaningful for comparing the relative interfacial behaviour of different systems. The values 2.5%, 5%, 7.5% and 10% represent a gradient ranging from conventional to high levels of incorporation. They represent a gradient in the mass fraction of the modifier. They are primarily used to simulate changes in mechanisms under polymer modification conditions and do not simply correspond to a single engineering solution. These incorporation levels are set for research purposes to cover the range of conventional and high incorporation levels that may occur in actual engineering applications, and to investigate performance characteristics. The model used in this study is of limited scale in order to balance computational efficiency with mechanistic interpretability; however, the model size is sufficient to compare relative changes under different loading levels and temperature conditions. As simulation duration affects the stability of diffusion curves and adhesion statistics, a pre-equilibrium–production phase simulation method was adopted to reduce random fluctuations. This study did not involve any explicit hydroxylation treatment of the aggregate surfaces. This was because the focus of this research was to compare relative interfacial behaviour under different modifier and dosage conditions; to ensure comparability across different operating conditions, representative mineral surfaces with simplified structures were employed. The effects associated with hydroxylation will be discussed further in subsequent research. To ensure that subsequent results—such as adhesion energy, relative concentration distribution and MSD—are based on a stable system, this study further conducted a convergence analysis of the time-dependent changes in potential energy, total energy, pressure and density during the MD simulation, as shown in Figure 7 and Figure 8. The results indicate that, in the early stages of the simulation, all parameters exhibited fluctuations of a certain magnitude. This was due to the system needing to gradually eliminate the internal stresses arising from the initial configuration and to achieve molecular rearrangement. As the simulation time increased, the potential energy and total energy gradually decreased and converged to a stable plateau, indicating that the internal energy of the system was continuously being released and eventually reached a state of equilibrium; simultaneously, following initial adjustments, the pressure and density also gradually stabilised within a narrow range of fluctuations, suggesting that the system’s volume and thermodynamic state had stabilised. It can be concluded from the above results that the system completed the main equilibrium process within the first 100 ps, after which all thermodynamic parameters remained within a stable fluctuation range. The box dimensions exhibit only limited fluctuations after reaching equilibrium, with no significant volume drift observed; this indicates that volume changes under the NPT ensemble will not have a significant impact on subsequent statistical analyses. Consequently, this paper utilises the post-equilibrium trajectories as production-phase data for subsequent analyses of interface structure and energy, thereby ensuring the reliability and comparability of the results. 3. Results 3.1. Qualitative Analysis of the Interaction Between Modified Asphalt and Aggregates For RDF, relative concentration distributions, and MSD analyses, only the equilibrated part of the trajectory was sampled. The instantaneous coordinates were directly used for structural statistics, while the box fluctuations were monitored simultaneously and found to remain within a narrow range after equilibration. Therefore, the volume variation under the NPT ensemble did not alter the qualitative trends of the post-processing results. Longitudinal sections of the final frame following the completion of MD simulations for the three modified asphalt–aggregate layer models are shown in Figure 9. It can be observed that the molecules of the modified asphalt and the aggregate are in close contact, and the modified asphalt is uniformly distributed across the aggregate surface. This indicates that the interaction between the modified asphalt and the aggregate layer model was highly effective during the simulation process; consequently, it is reasonable to use this model to investigate the diffusion, wetting and adhesion behaviour of the modified asphalt on the aggregate surface within this process. Temperature and the dosage of the modifier directly influence the interaction behaviour between modified asphalt and mineral aggregate at the interface. Figure 10 shows cross-sectional views of the Layer model at the interface between SBS-modified asphalt (with a modifier dosage of 5%) and mineral aggregate, obtained at different temperatures. Asphaltene, resin, aromatic and saturated fractions are marked in red, green, yellow and blue respectively. As can be seen from Figure 8, temperature directly affects the compactness of modified asphalt; the molecules of modified asphalt are less active at low temperatures and more active at high temperatures. Consequently, at low temperatures, the surface area of modified asphalt in contact with the aggregate is smaller, while as the temperature rises, the surface area of asphalt distributed across the aggregate increases. Furthermore, at low temperatures, the proportion of asphaltenes and colloids distributed on the aggregate surface is higher, whereas at high temperatures, the proportion of aromatic and saturated components distributed on the aggregate surface is greater. This also corroborates the widely observed finding in macroscopic tests that the mechanical behaviour of the asphalt–aggregate interface tends towards a brittle state at low temperatures and toward a viscoelastic state at high temperatures. Figure 11 shows cross-sectional views of the Layer model at the interface between modified asphalt and mineral aggregate at −20 °C and 160 °C. It can be observed that, regardless of whether conditions are at low or high temperatures, when the modifier content is low, the interface between the mineral aggregate and the modified asphalt is predominantly composed of light fractions of asphalt, such as aromatic and saturated fractions, while the proportions of gum and asphaltene are relatively low. However, as the modifier content increases, the proportion of aromatic and saturated fractions decreases, while the content of asphaltenes and resins rises. When the SBS content reaches 7.5%, a small number of SBS modifier molecules even appear at the interface. This indicates that during MD simulations, when the modifier content in the system is low, the aggregate is more prone to adsorbing the lighter components of the asphalt, facilitating smoother diffusion and wetting of the asphalt on the aggregate surface. This promotes the uniform distribution of asphalt on the aggregate surface but is not conducive to the formation of a contact surface with higher bonding strength at the interface. Conversely, the addition of the modifier exerts a certain locking effect on the asphalt, preventing the light fractions from diffusing to the aggregate surface. While this hinders the mixing of the modified asphaltic mixture, it enhances the strength at the asphalt–aggregate interface and improves the low-temperature performance of the mixture. Since the apparent phenomena observed in the MD simulations of SBR and PE are similar to those of SBS, we will not conduct a qualitative analysis of these two materials, but will instead proceed directly to a quantitative discussion of their wetting, diffusion and adhesion behaviour. 3.2. Quantitative Analysis of the Interaction Between Modified Asphalt and Aggregates 3.2.1. Wetting Effect of Modified Asphalt on Aggregates The relative concentration distribution of modified asphalt reflects its wetting effect on the surface of aggregates; therefore, the relative distribution concentration of modified asphalt on the surface of aggregates is shown in Figure 12 and Figure 13. As can be seen from Figure 12, the distribution concentration of the three types of modified asphalt on the aggregate surface is relatively uniform, showing a pattern of higher concentration in the centre and lower concentrations at the edges, indicating that all three types of modified asphalt exhibit excellent wetting effects on the aggregate. Although a visual analysis of the modified asphalt–aggregate interface reveals that changes in temperature alter the composition of the asphalt components distributed on the aggregate surface, the relative concentration distribution of the modified asphalt on the aggregate surface at different temperatures indicates that temperature does not alter the wetting area of the modified asphalt on the aggregate. Figure 13 illustrates the spatial distribution of the modified asphalt system on the aggregate surface. As can be seen from Figure 13, changes in the modifier dosage directly alter the concentration distribution of modified asphalt on the aggregate surface. At low dosage levels, there is little variation in the relative concentration distribution of modified asphalt across different positions on the aggregate surface; however, when the modifier dosage rises to 7.5%, the concentration of modified asphalt in the central part of the aggregate increases sharply, while the content at the edges decreases. This indicates that an increase in the modified asphalt content directly leads to uneven wetting on the aggregate surface, which indirectly confirms that high modifier content causes difficulties in mixing modified asphaltic mixtures. Among the three modifiers, this phenomenon is more pronounced with SBS and SBR, while it is less severe with PE. The non-uniform wetting observed at high modifier contents indicates that the spreading and coating ability of modified asphalt on the aggregate surface is constrained, which may lead to an uneven asphalt film thickness and locally insufficient coating or weak interfacial adhesion. Such a heterogeneous interfacial state can adversely affect mixture homogeneity and compaction quality, and may further induce local debonding, interfacial damage, and reduced durability during service. Therefore, although increasing modifier content can improve low-temperature interfacial strength, its negative influence on wetting uniformity and coating quality should also be considered in mixture design. 3.2.2. Diffusion Behaviour of Modified Asphalt in Aggregates Although relative distribution concentration provides a relatively intuitive representation of the distribution of modified asphalt on the surface of aggregates, it does not allow for an analysis of the entire process of diffusion of modified asphalt onto the aggregate surface. In contrast, mean square displacement (MSD) provides an excellent description of the diffusion process of particles within the system in the time domain; therefore, MSD is used to evaluate the diffusion behaviour of modified asphalt in aggregate. The calculation method for MSD is shown in Equation (1). The MSD values for the diffusion of the three types of modified asphalt within the system are shown in Figure 12 and Figure 13. M S D ( t ) = ⟨ | r i ( t ) − r i ( 0 ) | 2 ⟩ (3) where r i ( t ) is the position of particles in the system at each time, r i ( 0 ) is the initial position of the particles in the system and ⟨ ⟩ is the mean of all the particles in the system. As can be seen from Figure 14a,b, the diffusion process of modified asphalt on the aggregate surface can be divided into three stages. In the initial rapid diffusion stage, the asphalt molecules are still far from interfacial equilibrium and the driving force for migration toward the aggregate surface is strong; therefore, the molecular mobility is high and the interfacial voids are quickly reduced. In the subsequent steady-state diffusion stage, the asphalt molecules gradually form a relatively compact interfacial structure with the aggregate, and the diffusion rate becomes stable; this stage reflects the main period during which the binder wets and anchors onto the aggregate surface, and it is the most representative stage for evaluating interfacial diffusion behaviour. In the final reverse diffusion stage, the interfacial attraction approaches equilibrium and some molecules near the free surface or vacuum layer undergo slight rearrangement or backward migration, indicating that the system has nearly reached a stable thermodynamic state. Therefore, the three stages represent the transition from initial molecular migration to stable interfacial bonding, and finally to equilibrium reorganisation. It can be observed that during the steady-state diffusion phase, the higher the temperature, the higher the MSD value of the modified asphalt molecules, and the steeper the slope of the curve. This indicates that the diffusion depth and diffusion rate of modified asphalt on the aggregate surface are positively correlated with temperature; the higher the temperature, the faster the diffusion rate and the greater the diffusion depth. The diffusion coefficient of the modified asphalt molecules within the system can be calculated using Equation (2). D = 1 6 lim ∆ t → ∞ d M S D d ∆ t (4) As can be seen from Figure 12c, the higher the temperature, the higher the diffusion coefficient of the modified asphalt molecules, and the more active the molecular motion. At 180 °C, the diffusion rates of all three modified asphalt types are significantly higher than at 160 °C. Among these, SBR and PE exhibit a more pronounced increase in molecular diffusion rate with rising temperature, whereas SBS shows no significant increase. Although increasing the diffusion rate of modified asphalt at the mixing temperature is beneficial for the adhesion of asphalt to the aggregate surface, it also consumes more energy; therefore, an appropriate mixing temperature should be selected based on a balance between economic considerations and performance. When the temperature falls below 0 °C, the diffusion coefficient of modified asphalt molecules is less than half that at high temperatures. At this point, the movement of modified asphalt molecules on the aggregate surface becomes less active, and the asphalt–aggregate interface becomes harder and more brittle, which is detrimental to the low-temperature crack resistance of the asphaltic mixture. Furthermore, as the temperature drops further, the diffusion rate of modified asphalt molecules decreases even more, and the ductility of the asphalt film on the aggregate surface weakens, leading to a lack of toughness at the asphalt–aggregate interface. This provides a microscopic explanation for the cracking phenomenon observed in asphalt pavements in cold regions during actual engineering projects. As shown in Figure 15a,b, at low temperatures, an increase in the modifier content results in a slight decrease in the MSD of modified asphalt. This indicates that adding more modifier reduces the molecular activity of the modified asphalt to some extent at low temperatures, which theoretically reduces the toughness at the asphalt–aggregate interface. However, existing experiments generally indicate that high levels of modifier are beneficial for improving the low-temperature crack resistance of asphalt mixtures. In conjunction with the apparent phenomena shown in Figure 9, it can be observed that an increase in modifier content enhances the locking effect of the lighter fractions in modified asphalt. Combined with the higher strength and ductility of polymer modifiers, this results in improved bonding capacity at the asphalt–aggregate interface. This explains, from a microscopic perspective, why high levels of polymer modifier promote stronger low-temperature crack resistance in modified asphalt mixtures. As shown in Figure 15c,d, under high-temperature conditions, an increase in modifier content causes the MSD value of the modified asphalt to decrease significantly; not only does the diffusion depth become shallower, but the diffusion rate also decreases. Among these, the SBS modifier is the most sensitive, while PE is the least sensitive. This indicates that an increase in the modifier content significantly reduces the molecular activity of modified asphalt at high temperatures, and the greater the molecular weight of the modifier, the greater the effect. At mixing temperatures, modified asphalt exhibits a viscoelastic flow state; the higher the molecular activity, the lower the viscosity, and the better the mixing effect. The reduction in MSD and diffusion coefficient caused by high modifier content demonstrates, from a microscopic perspective, that an increase in modifier content directly leads to stronger interactions at the asphalt–aggregate interface. This hinders the uniform coating of modified asphalt onto the aggregates, thereby adversely affecting the road performance of the asphalt mixture. The diffusion rate of modified asphalt is closely related to pavement performance. A higher diffusion rate generally means that asphalt molecules can spread and wet the aggregate surface more rapidly, which is beneficial for improving coating efficiency, interfacial continuity and mixture homogeneity during the mixing stage. In contrast, a lower diffusion rate indicates stronger molecular restriction and higher system viscosity, which may increase the difficulty of aggregate coating and reduce construction workability. However, for long-term pavement performance, a moderate reduction in the diffusion rate at low temperatures may also reflect stronger interfacial constraint and improved resistance to deformation and cracking. 3.2.3. Adhesion Between Modified Asphalt and Aggregates The adhesion energy between modified asphalt and mineral aggregate directly characterises the adhesion properties between the two. This is calculated as the difference in energy within the system before and after the MD simulation, as shown in Equation (3), and can be classified into electrostatic forces and van der Waals forces depending on the source of the energy. The calculation results for modified asphalt and mineral aggregate are presented in Table 6. ∆ E = E A + E B − E A B (5) where E A and E B are the energies when A and B molecules exist alone, respectively; E A B is the energy when A and B coexist; ∆ E is the adhesion work between A and B. When ∆ E is greater than 0, it represents mutual adsorption. When ∆ E is less than 0, it represents mutual repulsion. As shown in Figure 16, the adhesion energy between modified asphalt and aggregate is higher at low temperatures and lower at mixing temperatures, and is primarily influenced by van der Waals forces. At low temperatures, the adhesion between modified asphalt and aggregate is better, whereas at high temperatures, the adhesion is weaker. Temperature directly affects the degree of molecular activity in modified asphalt; the higher the temperature, the greater the activity, but this results in reduced adhesion between the modified asphalt and the aggregate. From a mechanistic perspective, the higher the adhesion energy shown in Figure 16, the stronger the interaction between the modified asphalt molecules and the aggregate surface, and the more the system tends to form a stable adsorbed layer; this is consistent with the more concentrated interfacial distribution at low temperatures shown in Figure 12 and Figure 13, and the phenomenon of uneven wetting caused by variations in the admixture content. Therefore, Figure 16 is not an isolated energy result, but rather an important basis for explaining the adsorption strength on the aggregate surface, interfacial stability and wetting behaviour. The increase in interfacial adhesion energy at low temperatures indicates that modified asphalt and aggregate can form a more stable adsorptive bond, which helps improve interfacial tensile strength and resistance to stripping, thereby reducing the likelihood of crack initiation and propagation under cold conditions. For pavements in cold regions, a higher low-temperature adhesion energy means a more robust binder–aggregate interface, which can delay interfacial damage development under combined traffic loading and thermal contraction, and thus improve long-term pavement durability. Meanwhile, the fact that adhesion is mainly governed by van der Waals forces suggests that the interfacial interaction at low temperatures is dominated by physical adsorption; therefore, increasing modifier content to strengthen such interactions can enhance crack resistance, while still requiring a balance with high-temperature workability and coating performance. The addition of a modifier increases the work done by the van der Waals forces between the modified asphalt and the aggregate, while reducing the work done by electrostatic forces, resulting in an overall increase in the adhesion energy. This indicates that the addition of a modifier enhances the intermolecular forces between the polymer and silica within the system, and this effect becomes more pronounced as the modifier content increases. In low-temperature conditions, increasing the modifier content can enhance the adhesion energy between modified asphalt and mineral aggregate, thereby promoting stronger adhesion between aggregate particles and reducing the likelihood of cracking in the asphaltic mixture. However, as the modifier content increases, the viscoelastic work between the modified asphalt and the aggregate at the mixing temperature becomes greater, reducing the workability of the asphalt mixture. This makes it more likely for white patches to form in the modified asphalt mixture, ultimately compromising the road performance of the asphalt mixture. In summary, temperature and the dosage of the modifier directly influence the wetting, diffusion and adhesion of modified asphalt on the surface of aggregate. Increasing the dosage of modifier helps to improve the strength and toughness of the modified asphalt–aggregate interface at low temperatures, but at high temperatures it affects the mixability of the mixture. A comparison of the three modifiers reveals that SBS and SBR are more sensitive to changes in temperature and dosage; at high dosages, they are more likely to cause diffusion limitations and uneven interfacial wetting. PE exhibits a relatively slower rate of diffusion decay, but at high dosages it similarly increases interfacial constraints and affects the uniformity of encapsulation. Overall, SBS performs best in terms of low-temperature interface reinforcement, followed by SBR, while PE offers greater advantages in maintaining a certain degree of flowability; however, all three require a balance to be struck between low-temperature crack resistance and high-temperature workability. As shown in Table 7, in order to summarise the general patterns of modified asphalt interface behaviour under different temperature conditions and at different modifier dosages, this paper further summarises the characteristics of wetting, diffusion and adhesion, along with their engineering implications. This summary table facilitates a comprehensive comparison of the differences in response between the three systems—SBS, SBR and PE—under service and construction conditions in cold regions. To evaluate the robustness of the simulation results, a sensitivity analysis was conducted with respect to key parameters, including temperature, modifier content, and simulation duration. The results indicate that temperature has a significant effect on interfacial diffusion behaviour, with higher temperatures leading to enhanced molecular mobility and increased diffusion capacity. In contrast, increasing modifier content suppresses molecular migration and tends to aggravate non-uniform wetting at elevated temperatures. Furthermore, consistent trends observed across different temperature and dosage conditions demonstrate the stability of the model response. Regarding simulation duration, the division into equilibration and production stages ensures that all statistical data are obtained from equilibrated systems, thereby minimising the influence of initial configurations and transient fluctuations. Overall, the results exhibit good sensitivity and reliability within the investigated parameter range. The methodology employed in this study has certain limitations, namely that it has not yet been experimentally validated; furthermore, the asphalt molecular model is a simplified representative model comprising four components and twelve molecules, and the simulation scale differs from that of real-world engineering applications. Future research will combine experiments with simulations on a larger scale to further validate the findings. This study employed a simplified model of representative mineral surfaces, without explicitly accounting for the hydroxylation of aggregate surfaces, the multi-phase composition of minerals, or the complex interfacial chemical characteristics of alkaline aggregates. Consequently, the conclusions drawn in this paper are primarily applicable to the analysis of general trends on representative mineral surfaces. In future work, models of hydroxylated surfaces and aggregates with different mineral compositions will be developed to assess the impact of surface chemical differences on the adsorption and adhesion behaviour of modified asphalt. Although the aggregate surface was hydroxylated, the hydroxyl density and spatial distribution were still simplified in the present model. Future work will consider different hydroxylation degrees and mineral compositions to further improve realism. 4. Conclusions This paper argues that, when designing road pavements for cold regions, priority should be given to comparing different polymer systems—SBS, SPR and PE—in order to select the most suitable material; while high polymer content enhances low-temperature performance, it can impair mix workability; furthermore, in terms of construction strategies, mixing and compaction processes should be optimised in conjunction with the temperature window. This paper employs MD simulation to analyse the mechanisms of microscopic diffusion and adhesion at the interface between polymer-modified asphalt and aggregate under high polymer content. A molecular interaction model between modified asphalt and aggregate was established to investigate the interaction effects between modified asphalt and aggregate under low- and high-temperature conditions, leading to the following conclusions. (1) Temperature directly influences the reactivity of modified asphalt molecules; they are inactive at low temperatures and active at high temperatures. At low temperatures, a higher proportion of asphaltenes and colloids are distributed on the aggregate surface, whereas at high temperatures, a greater proportion of aromatic and saturated fractions are distributed on the aggregate surface. (2) At low temperatures, the total adhesion energies of SBS, SBR and PE increased from 980.825, 628.180 and 736.990 kcal/mol at 2.5% to 1469.025, 1030.982 and 818.564 kcal/mol at 10%, representing increases of approximately 49.8%, 64.1% and 11.1% respectively. This indicates that increasing the blend ratio enhances interfacial adhesion at low temperatures, which is beneficial for improving crack resistance. (3) Whether at low or mixed temperatures, SBS exhibited the highest overall adhesion energy, indicating that it has the strongest reinforcing effect on the aggregate interface; SBR ranks second, whilst PE is relatively weaker. Notably, at a 10% content, the total adhesion energy of SBS reaches 1469.025 and 672.120 kcal/mol at low and high temperatures respectively, both higher than the other two modifiers, indicating that SBS is more suitable for pavement systems in cold regions where both low-temperature crack resistance and interface reinforcement are required. (4) When the admixture content is increased to 7.5% or higher, the concentration in the central region of the aggregate surface rises significantly, whilst the concentration in the peripheral regions decreases, indicating that high admixture content leads to uneven wetting; high admixture content reduces the MSD and diffusion coefficient at high temperatures, resulting in a decrease in the spreading ability of the modified asphalt on the aggregate surface. Consequently, although high-content modification helps to enhance low-temperature interfacial adhesion, it gives rise to issues such as mixing difficulties, uneven coating and reduced workability. (5) This paper compares the differences in response of three types of polymer-modified systems—SBS, SBR and PE—under varying temperatures and loading levels. It identifies differences in the sensitivity of various modifiers to diffusion, wetting and adhesion, and concludes that high-loading modifiers exert a dual effect on low-temperature toughness and high-temperature workability. (6) Temperature and modifier dosage jointly influence the diffusion and wetting behaviour of modified asphalt on aggregate surfaces; while increased temperature promotes interfacial spreading and encapsulation efficiency, high dosage inhibits molecular migration and leads to uneven wetting at high temperatures, thereby increasing sensitivity to mixing conditions during the construction phase. Interfacial diffusion capacity is closely related to the uniformity of the mixture, while interfacial adhesion is significantly enhanced at low temperatures, helping to improve interfacial bond strength and inhibit crack propagation. Author Contributions Conceptualization, W.Y. and S.Z.; methodology, X.B.; validation, X.W. and J.Y.; formal analysis, C.Z.; investigation, X.W.; resources, W.Y.; data curation, W.Y.; writing—original draft preparation, W.Y.; writing—review and editing, W.Y. and C.Z.; visualisation, S.Z.; supervision, C.Z.; project administration, S.Z.; funding acquisition, W.Y. All authors have read and agreed to the published version of the manuscript. Funding This research was funded by China Construction 7th Engineering Bureau grant number [JTZB-YZKS-D081/2024]. And The APC was funded by China Construction 7th Engineering Bureau. Data Availability Statement The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author. Conflicts of Interest Author Wei Yuan, Shaobo Zhang, Xiaohui Bu, Xudong Wang and Jiahao Yang were employed by the company Communications Construction Company of CESEC 7th Division Corp. Ltd, Zhengzhou, China. The remaining author declared that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Abbreviations The following abbreviations are used in this manuscript: SBS Styrene–butadiene–styrene SBR Styrene–Butadiene Rubber PE Polyethylene MD Molecular Dynamics MSD Mean Square Displacement NPT Isothermal–isobaric ensemble References Chen, S.; Zheng, C.; Chen, W.; Guo, W.; Wei, Z.; Li, H.; Xu, J. Study on viscoelastic properties and rheological properties of SBS modified asphalt for road use with service life. Int. J. Adhes. Adhes. 2025, 136, 103870. [ Google Scholar] [ CrossRef] Lin, M.; Lei, Y.; Li, P.; Wang, Z.L. Study on the Rheological Properties and Modification Mechanism of Graphene/Rubber Composite-Modified Asphalt. Adv. Civ. Eng. Mater. 2025, 14, 1–22. [ Google Scholar] [ CrossRef] Zhao, C.; Li, Y.; Li, R.; Li, X.; Lyu, L.; Pei, J. Study on the pavement performances and the modification mechanism of dry-modified asphalt with rubber-plastic composite particles. Environ. Technol. Innov. 2025, 37, 103986. [ Google Scholar] [ CrossRef] Wu, M.; You, Z. Molecular dynamics models to investigate the diffusion behavior of emulsified asphalt on the aggregate surface. Constr. Build. Mater. 2023, 409, 134061. [ Google Scholar] [ CrossRef] Jin, J.; Chen, H.; Liu, S.; Xiao, M.; Liu, L. Study on preparation and properties of phase change modified asphalt for the functional pavement. Constr. Build. Mater. 2024, 439, 137248. [ Google Scholar] [ CrossRef] Wang, Y.; Guo, S.; Pei, Z.; Zhan, S.; Lin, S.; Ma, K.; Lei, J.; Yi, J. Study of the Properties and Modification Mechanism of SBS-Modified Asphalt by Dry Process. Materials 2024, 17, 1454. [ Google Scholar] [ CrossRef] Shao, L.; Gao, Y.; Zhang, X.; Lv, X.; Wang, Y. Self-Healing Properties of Graphene Oxide-Modified Asphalt Based on Molecular Dynamics. J. Mater. Civ. Eng. 2025, 37, 04024514. [ Google Scholar] [ CrossRef] Zhao, S.; Ouyang, J.; Zhou, L. Rheological Properties of Polyurethane-Modified Asphalt and Its Effect on the Performance of Porous Asphalt Mixes. J. Mater. Civ. Eng. 2025, 37, 04025327. [ Google Scholar] [ CrossRef] Chen, W.; Zhang, Z.; Wei, J.; Zhang, X.; Gan, C.; Wang, W.; Sun, Y. Research on Mechanical Performance of Porous Asphalt Mixture with High-Viscosity Modified Asphalt. Appl. Sci. 2025, 15, 3631. [ Google Scholar] [ CrossRef] Chen, S.; Zhang, B.; Zheng, H.; He, X.; Su, Y.; Xu, H.; Zhu, Y.; Pan, Y.; Chen, J. Polymerized activating highly-dispersible waste tire powder for viscoelasticity regulation of modified asphalt at a lower mixing temperature. Constr. Build. Mater. 2023, 366, 130216. [ Google Scholar] [ CrossRef] Song, Z.; Wei, Z.; Zheng, C.; Li, H.; Zhao, J.; Luo, H.; Jin, W.; Wang, F. Microscopic mechanism and effect analysis of polymer modifiers on embrittlement and viscosity behaviour of asphalt. Results. Eng. 2024, 22, 102204. [ Google Scholar] [ CrossRef] Greenfield, M.L. Molecular modelling and simulation of asphaltenes and bituminous materials. Int. J. Pavement Eng. 2011, 12, 325–341. [ Google Scholar] [ CrossRef] Wang, C.; Huang, S.; Chen, Q.; Ji, X.; Duan, K. Materials, preparation, performances and mechanism of polyurethane modified asphalt and its mixture: A systematic review. J. Road Eng. 2023, 3, 16–34. [ Google Scholar] [ CrossRef] Huang, T.; Lin, B.; Cao, Z.; Yi, J.; Yu, X.; Qian, G. Laboratory testing of road performance and mechanism study of SBR/APAO composite modified asphalt and its mixtures. Case Stud. Constr. Mater. 2025, 22, e04525. [ Google Scholar] [ CrossRef] Dong, F.; Lu, J.; Yu, X.; Yang, P.; Jin, Y.; Wan, L.; Ding, Y.; Chen, H. Insight into the diffusion behaviors of SBS-compatibilizer-coupled modified asphalt: Molecular dynamics simulation and experimental validation. J. Vinyl Addit. Technol. 2024, 30, 1270–1284. [ Google Scholar] [ CrossRef] Chen, B.; Dong, F.; Yu, X.; Zheng, C. Evaluation of Properties and Micro-Characteristics of Waste Polyurethane/Styrene-Butadiene-Styrene Composite Modified Asphalt. Polymers 2021, 13, 2249. [ Google Scholar] [ CrossRef] Yan, K.; Shi, K.; Wang, M.; Huang, J.; Li, Y.; Ouyang, Z. Effect of high-density-polyethylene (HDPE) and waste tire rubber (WTR) on laboratory performance of styrene butadiene styrene (SBS) modified asphalt. Road Mater. Pavement Des. 2025, 26, 1360–1375. [ Google Scholar] [ CrossRef] Zhao, S.; Zhang, H.; Feng, Y.; Hang, Y. Effect of diffusion on interfacial properties of polyurethane-modified asphalt–aggregate using molecular dynamic simulation. Rev. Adv. Mater. Sci. 2022, 61, 778–794. [ Google Scholar] [ CrossRef] Zhang, H.; Gong, M.; Gao, D.; Yang, T.; Huang, Y. Comparative analysis of mechanical behavior of composite modified asphalt mixture based on PG technology. Constr. Build. Mater. 2020, 259, 119771. [ Google Scholar] [ CrossRef] Ma, H.; Guo, F.; Han, J.; Zhi, P. Analysis of Rheological Properties and Regeneration Mechanism of Recycled Styrene–Butadiene–Styrene Block Copolymer (SBS) Modified Asphalt Binder Using Different Rejuvenators. Materials 2024, 17, 4258. [ Google Scholar] [ Cros
