The Effect of Aging on the Molecular Distribution of Crumb Rubber Modified Asphalt Based on the Gel Permeation Chromatography Test

Abstract

Asphalt aging is one of the main causes of asphalt pavement cracking, loosening and other issues. On a macro level, the asphalt hardens and becomes more brittle, while on a micro level, the chemical composition and molecular weight distribution change. This is a very complicated physicochemical process. Gel permeation chromatography (GPC) is a powerful technical tool for understanding the mechanism of asphalt aging and expressing the asphalt aging process. GPC can be used to measure the distribution and content of each component in the asphalt aging process. The mechanism of action of crumb rubber modified asphalt has not been fully elucidated due to its complex composition. This study investigated the molecular weight of crumb rubber modified asphalt before and after aging, and filtered asphalt based on gel permeation chromatography. The results indicated that crumb rubber itself experiences severe degradation following PAV aging and that a significant number of macromolecular materials are incorporated into the asphalt phase, causing changes in key parameters. The average molecular weight (M_w) and dispersion (D) of crumb rubber modified asphalt are directly related. At the same time, the M_w of crumb rubber modified asphalt has a positive correlation with LMS content, while SMS content has a negative correlation with M_w. The increase in crumb rubber content has a positive impact on the material interaction, and the molecular weight distribution of crumb rubber modified asphalt is affected by the reaction degree of the crumb rubber in asphalt. The complex physical and chemical reaction of crumb rubber in asphalt has a direct impact on the external macro rheological properties of asphalt.

1. Introduction

Rubber asphalt is a type of modified asphalt binder material created by processing waste tires into crumb rubber particles and mixing them thoroughly with base asphalt at high temperatures. It is the ideal environmentally friendly pavement material and has emerged as a focus of research in the area of materials for highway engineering [1]. The study of rubber asphalt’s impact on pavement performance both domestically and internationally is currently at a fairly advanced stage. The advancement of rubber asphalt technology for use in road engineering is the main emphasis of research on rubber asphalt [2]. The most typical approach of characterizing rubber asphalt is primarily using conventional mechanical indicators to assess its high-temperature rheological qualities and low-temperature cracking properties. At present, the main emphasis is on the rubber asphalt performance control and mechanism. Through an in-depth understanding of the modification mechanism of rubber asphalt, the manufacture of rubber asphalt may be efficiently optimized, and the road performance of rubber asphalt can be significantly enhanced [3,4]. Due to the fact that asphalt is a mixture of various compounds, and tire crumb rubber is composed of natural rubber or synthetic rubber, sulfur, carbon black, plasticizer and antioxidant, which increases the complexity of rubber asphalt components, there must be some limitations to only analyzing the modification mechanism of rubber asphalt using macroscopic test indicators. In light of this, it is essential to comprehend the diverse microstructure properties of rubber asphalt and be able to effectively describe rubber asphalt at the micro size; therefore, a scientific method for rubber asphalt study and improvement has become unavoidable [5]. In terms of studies on the application mechanism of rubber asphalt, researchers agree that there are complex physical and chemical behaviors in the preparation process of rubber asphalt, and this complex behavior is affected by the chemical composition and microstructure of the base asphalt and rubber particles. At the same time, rubber asphalt is composed of chemical molecules with different molecular weights, and the complex molecular morphology also affects the chemical properties, rheological properties and microstructure of rubber asphalt, thus significantly affecting its intrinsic properties [6,7].

Gel permeation chromatography (GPC) is a type of liquid chromatography, which can be used to determine the relative molecular weight distribution and molecular size of asphalt. The macroscopic properties of asphalt are closely related to its relative molecular weight. The larger the molecular weight, the stronger the intermolecular force. After asphalt aging, its molecular weight increases, resulting in an increase in asphalt viscosity, an increase in its softening point, a decrease in penetration and a decrease in temperature sensitivity [8,9]. Through GPC, it has been found that the dispersion coefficient D of SBR (styrene butadiene rubber) modified asphalt’s molecular weight increases, which is believed to promote its low-temperature performance [10,11]. Herrington et al. used gel permeation chromatography to study the changes in the molecular weight distribution of asphalt before and after aging, and found that the dispersity and average molecular weight of asphalt increased significantly during aging. The distribution and molecular weight of aged asphalt have been extensively examined by scientists. Moreover, despite the fact that the results vary, it is commonly accepted that the content of high-molecular-weight compounds has increased, i.e., as asphalt ages, its asphaltene content increases and its molecular weight increases [12]. Many researchers have applied GPC to asphalt chemistry research, mainly studying polymer modified asphalt and the aging mechanism of asphalt [13]. It is also of great significance to apply research to the field of rubber asphalt in order to study the modification mechanism of rubber asphalt. With the aid of gel chromatography, which separates and quantifies the rubber in modified asphalt, the rubber content of rubber asphalt may be reliably ascertained. For practical application, this method’s great accuracy and precision are ideal [14,15]. Additionally, gel chromatography can be used to determine the number and distribution of macromolecules in rubber asphalt. This allows researchers to study and analyze the swelling and depolymerization process of crumb rubber in asphalt, which is crucial for determining the best swelling method for crumb rubber [16,17]. The viscosity and modification mechanism of rubber asphalt were studied by gel permeation chromatography (GPC). The results showed that the macromolecular content (LMS) of rubber asphalt increased significantly compared with base asphalt [18,19]. Baek et al. also highlighted that the molecular weight and content of rubber asphalt affect the rheology of rubber asphalt, in which the LMS index and rutting factor G */sin δ have a strong correlation. Therefore, G*/sin δ of rubber asphalt can be predicted based on the LMS value, the crumb rubber content and the particle size [20]. Shen et al. used gel permeation chromatography to test the change pattern of rubber particles in the dry process. The results indicate that the percentage of large molecular weight (LMS) in rubber asphalt increased with the extension of the high-temperature storage time of the mixture, and the mixture performance gradually became close to the effect of rubber asphalt in a wet process with the extension of storage time [10,21]. Peramanu analyzed asphalt from two sources, mainly investigating the molecular size distribution (MSD) of four components of asphalt. The research results were consistent with the existing understanding that the molecular weight of the light component of asphalt is smaller than that of the heavy component. The aging of asphalt causes a change in the asphalt MSD; therefore, the aging state of asphalt can be studied by GPC [22]. Garrick and Wood’s research shows that the percentage of LMS in asphalt will increase after heating. Kim et al. evaluated the change in the LMS percentage of asphalt after short-term aging, and the research results showed that the percentage of LMS increased after the short-term aging of asphalt. In addition to the base asphalt, the results of GPC chromatogram analysis of asphalt before and after the aging of rubber modified asphalt led to the same conclusion [23]. Similarly, the aging degree of asphalt in different graded mixtures is also characterized by GPC. Lee et al. studied the influence of rubber particles on asphalt using GPC. In addition, GPC has also been used to determine the prediction results of molecular dynamic models [24]. Ding et al. used GPC to verify the accuracy of the molecular dynamic model. Existing studies have mainly focused on the perspective that “the molecular weight of different components in asphalt is different”, and characterized the changes in asphalt or detected the content of polymer organic modifier in asphalt by comparing the change in asphalt MSD. In addition, some scholars have used GPC to detect asphalt due to its high sensitivity [25,26].

The GPC method can help us to understand the mechanism of reaction of modified asphalt and can be used to quantitatively describe the aging condition of asphalt. Regarding crumb rubber modified asphalt, by describing the change in the molecular weight and its components, the influence of crumb rubber on the distribution and molecular weight of its components can be effectively characterized, thus providing support for building the relationship between crumb rubber modified asphalt and macro performance. Based on the GPC method, this study analyzed the molecular weight of original asphalt, aged asphalt and filtered asphalt, and analyzed the influence of the aging effect and crumb rubber on the change in the molecular weight of asphalt. Furthermore, the relationship between the interaction effect and filling effect was determined, and the immediate result of the crumb rubber modified asphalt material exchange reaction was obtained, providing a research direction for the mechanism of crumb rubber modified asphalt.

2. Materials and Methods

2.1. Materials

2.1.1. Asphalt Binder

This study used two sources and three grades of base asphalt to create crumb rubber modified asphalt by mixing crumb rubber with base asphalt in order to obtain test results that were easily distinguishable. According to the Chinese specification known as the “Test Specification for Asphalt and Asphalt Mixture in Highway Engineering” (JTG E20-2011), the fundamental characteristics of the SK70#, JB70# and JB90# asphalt used in this study were tested. The exact performance findings are presented in

Table 1. Performance results of base binder.

Test Items	SK70#	JB70#	JB90#	Quality Index	Test Method
Penetration (25 °C, 5 s, 100 g)/0.1 mm	71	71.2	94	60–80	T0604
Softening point (R&B)/°C	46.8	46.3	46.4	≥46	T0606
60 °C Dynamic viscosity/Pa.s	186	192	185	≥180	T0620
10°C Ductility/cm	38	33	46	≥20	T0605
RTFOT (163 °C, 5 h)
Quality change/%	0.14	0.18	0.24	≤0.6	T0609
Residual penetration ratio (25 °C)/%	67	76	89	≥65	T0604
Residual ductility (10 °C)/cm	6.7	7.6	15.5	≥6	T0605

.

2.1.2. Crumb Rubber

Basic Properties of Crumb Rubber

Crumb rubber is a type of powdered rubber material processed from waste products. Its main components include synthetic rubber, natural rubber, plasticizer, carbon black and ash. This study focused on the use of crumb rubber pavement modification to increase asphalt pavement’s low-temperature crack resistance and driving comfort. From the perspective of adding crumb rubber, this study used 30, 40 and 80 mesh crumb rubber produced by the same plant, and its main technical indicators are shown in

Table 2. Main performance indices of crumb rubber.

Technical Indicators	Measured Results	Technical Standard	Test Method
Residue/%	6.5	<10	GB/T 19208
Relative density/kg/m3	1.13	1.10~1.30	GB/T 19208
Water content/%	0.46	<1	GB/T 19208
Metal content/%	0.007	<0.05	GB/T 19208
Fiber content/%	0.06	<1	GB/T 19208
Natural rubber content/%	34	≥30	GB/T 13249-91
Ash content/%	7	≤8	GB 4498-1997
Acetone extract/%	6	≤22	GB/T 3516
Carbon black content/%	29	≥28	GB/T 14837
Rubber hydrocarbon content/%	58	≥42	GB/T 14837

.

2.1.3. Test Scheme

The test was undertaken using the following method: The base asphalt was preheated to 177 °C, and 30, 40 and 80 mesh crumb rubber was selected, produced by the same manufacturer; then, the external mixing method was adopted, producing a crumb rubber content of 10%, 15% and 20%, respectively. Mixing was carried out at a low speed by using a mechanical mixer and the crumb rubber was added while mixing. After all of the crumb rubber was added into the asphalt, the speed of the mixer was controlled at 1000 rpm and the temperature was 177 °C. Mixing was carried out for 30 min in order to prepare the crumb rubber modified asphalt.

2.2. Aging Test

The asphalt aging test was used to simulate the aging behavior of asphalt during both the construction and post-construction phases. It was divided into a short-term aging test and long-term aging test according to the aging stage.

The indoor short-term aging of asphalt mainly simulates the thermal oxygen behavior of asphalt due to heating during storage, transportation, mixing and paving. The rolling thin film oven test (RTFOT) was used for the test. An amount of 35 ± 0.5 g asphalt was placed into a sample bottle and put it into a rotating film oven to heat at 163 ± 0.5 °C for not less than 75 min; the sample bottle rotated at a speed of 15 ± 0.2 r/min. The specific operation steps were carried out in accordance with the ASTM D2872 test method.

In this paper, Prentex 9300 PAV was used for the long-term aging of asphalt. An amount of 50 ± 0.5 g of RTFOT asphalt sample was placed into the sample tray. The sample tray was then put into the PAV, maintaining the air pressure inside the PAV at 2.1 ± 0.1 MPa; the temperature was kept at 100 °C, and aging continued for 20 h.

2.3. Filtering Test

The modification of asphalt by crumb rubber is based on both the interaction effect (IE) between the crumb rubber and asphalt and the particle effect (PE) of the crumb rubber particles in asphalt, as shown in Formulas (1) and (2). The final result of the rubber modification of asphalt is shown in Formula (3). In order to verify the influence of crumb rubber on the composition and structure of asphalt, the crumb rubber was filtered from the modified asphalt by extraction and filtration, and then relevant tests were carried out on the filtered crumb rubber and extracted asphalt separately to explore the actual role of crumb rubber in asphalt.

$$\mathrm{IE}=\frac{{\mathrm{I}}_{\mathrm{CRM}}-{\mathrm{I}}_{\mathrm{BB}}}{{\mathrm{I}}_{\mathrm{BB}}}$$

(1)

$$\mathrm{PE}=\frac{{\mathrm{I}}_{\mathrm{CRM}}-{\mathrm{I}}_{\mathrm{EB}}}{{\mathrm{I}}_{\mathrm{BB}}}$$

(2)

$$\mathrm{CRE}=\frac{{\mathrm{I}}_{\mathrm{CRM}}-{\mathrm{I}}_{\mathrm{BB}}}{{\mathrm{I}}_{\mathrm{BB}}}$$

(3)

2.4. Gel Permeation Chromatography (GPC)

In this study, a high-performance liquid chromatography (HPLC) system from Waters was used for gel permeation chromatography testing. PS (polystyrene) resin was used as the standard object, the pore diameter of the chromatographic column was 10 Å, and the particle size was 5 μm. The conditions were: column temperature 35 °C, column pressure 625PSI, solvent tetrahydrofuran, and flow rate 1 mL/min. A total of 20 mg of sample was weighed on a microbalance and dissolved in a 10 mL volumetric flask with tetrahydrofuran; the dissolution time of the crumb asphalt was 24 h, and the dissolution time of base asphalt and four components was 12 h. Once the instrument baseline leveling was complete, necessary steps to accomplish the test proceeded.

The molecular weight of asphalt and its constituent parts were quantitatively analyzed in this paper after the measured sample data were converted into molecular weight. This was greater than 19,000 for macromolecules (LMS), between 3000 and 19,000 for intermediate molecules (MMS) and less than 3000 for small molecules (SMS). Microsoft Excel was used to calculate and handle all of the data.

Through the fitting formula obtained above, the normalized data obtained from the GPC test were processed to obtain the relative molecular weight of the test sample. At the same time, the weight average molecular weight M_w, the number average molecular weight M_n, the Z average molecular weight and the dispersion D of each component were determined simultaneously using Formulas (4)–(7) [27].

$$\overline{M_w}=\frac{{\displaystyle \sum N_i{M}_i^2}}{{\displaystyle \sum N_i{M}_i}}=\frac{{\displaystyle \sum W_i{M}_i}}{{\displaystyle \sum W_i}}$$

(4)

$$\overline{M_n}=\frac{{\displaystyle \sum N_i{M}_i}}{{\displaystyle \sum N_i}}=\frac{{\displaystyle \sum W_i}}{{\displaystyle \sum {W_i/M}_i}}$$

(5)

$$\overline{M_Z}=\frac{{\displaystyle \sum N_i{M}_i^3}}{{\displaystyle \sum N_i{M}_i^2}}=\frac{{\displaystyle \sum W_i{M}_i^2}}{{\displaystyle \sum W_i{M}_i}}$$

(6)

$$D=\frac{\overline{M_w}}{\overline{M_n}}$$

(7)

where N_i is the number of molecules with a molecular weight of M_i; W_i is the weight of the component with a molecular weight M_i.

3. Results and Discussion

3.1. Base Asphalt GPC Analysis

Firstly, the molecular weight and distribution of SK70, JB70 and JB90 base asphalt before and after aging were studied. Figure 1 shows the average molecular weight distribution of SK70 asphalt samples after the original sample, as well as both short-term aging and long-term aging; the other two types of asphalt were basically the same. It can be seen from the results that the molecular weight distribution curve of asphalt exhibited obvious changes after aging, and its molecular weight moved in the macromolecular direction as a whole, indicating that the macromolecular components in asphalt significantly increased after aging. At the same time, on the molecular weight distribution curve of short-term aging and long-term aging, there was an obvious shoulder peak in the molecular weight distribution curve between 10⁴ and 10^4.2. According to the molecular weight division standard, this interval was the asphaltene distribution area, suggesting that aging leads to an intermolecular polymerization reaction in asphalt and promotes the increase in macromolecular content.

The M_w, M_n, M_z and D values of asphalt before and after molecular aging were obtained by processing the GPC data according to Formulas (4)–(7). At the same time, molecular weight division was divided into three parts according to the molecular weight division method adopted in this paper (See

Table 3. Changes in molecular weight of base asphalt before and after aging.

Name	Mw	Mn	Mz	Mp	D	LMS	MMS	SMS
SK70-OB	2574	853	5995	656	3.02	0.00	27.37	72.63
SK70-RTFO	2630	886	6219	807	2.97	0.15	27.58	72.27
SK70-PAV	3240	946	8260	837	3.43	0.85	31.83	67.32
JB70-OB	2742	847	7426	530	3.24	0.47	28.17	71.36
JB70-RTFO	3194	897	7960	505	3.56	0.68	32.28	67.04
JB70-PAV	3483	920	9043	484	3.79	1.27	33.66	65.08
JB90-OB	3044	875	15,046	520	3.48	0.81	30.23	68.96
JB90-RTFO	4761	934	67,487	524	5.10	2.07	33.59	64.35
JB90-PAV	5043	955	105,768	531	5.28	2.44	34.92	62.63

for specific results). After aging, M_w-, M_n-, M_z- and D-related indices all increased by various degrees. With regard to long-term aging, the M_w, M_n, M_z and D of SK70 asphalt increased by 25.9%, 10.9%, 37.8% and 13.5%, respectively, compared with the original asphalt; the corresponding indices of JB70 asphalt increased by 27.0%, 8.5%, 21.8% and 17.0%, respectively, whereas the corresponding indices of JB90 asphalt increased by 65.66%, 9.1%, 6.03% and 51.8%, respectively. The above results show that the increase in M_w and D for the three types of asphalt was relatively significant, while the increase in M_n was relatively stable, maintained between 8.5~10.9%. This demonstrates that as asphalt ages, its macromolecular content noticeably rises. This outcome is mostly the result of association, polymerization and other related reactions in the asphalt, which lead to a rise in the asphalt’s macromolecular component content. This conclusion can also be drawn from the tendency for change in the composition of each asphalt in terms of macromolecules, intermediate molecules and small molecules. The amount of macromolecules and intermediate molecules generally increased with increased aging, while the amount of small molecules steadily declined. There were some differences in the trends in the aforementioned indicators for the three base asphalts, notably JB90 asphalt’s M_w, D, LMS and SMS changes. These were somewhat greater than those of the other two asphalts and suggest that the JB90 asphalt has a weaker aging resistance. This is primarily because high-grade asphalt components contain a high amount of dispersion medium, which makes it simple for oxidation to convert non-functional group substances into esters. Asphalt is easily converted into a multi-molecular or amphiphilic component structure by functional groups, causing polar association reactions at various points in the asphalt and the subsequent production of a three-dimensional structured sub-microstructure. High temperatures cause these structures to connect or cleave in order to produce additional oxygen reactions.

3.2. Crumb Rubber Modified Asphalt GPC Analysis

Figure 2 shows the GPC results of SK70 crumb rubber modified asphalt. Compared with the original SK70 base asphalt, the molecular weight distribution range of the original crumb rubber modified asphalt and the aged PAV with a crumb rubber content of 10%, 15% and 20% was significantly larger than that of the base asphalt. This indicates that the molecular weight of the crumb rubber modified asphalt is larger than that of the base asphalt, indicating that the addition of crumb rubber will increase the molecular weight of the asphalt to varying degrees. At the same time, with the increase in the amount of crumb rubber, the molecular weight of the asphalt will increase, and this increase will be more obvious after PAV aging. This phenomenon indicates that the crumb rubber degrades after PAV aging, resulting in the increase in macromolecular content in the asphalt phase and the overall increase in macromolecular weight distribution. This result was consistent with previous research results [28,29]. In order to further quantitatively analyze the molecular weight change in crumb rubber modified asphalt, relevant information regarding M_w, M_n, M_z, D and the molecular weight distribution before and after aging can be obtained by processing GPC data.

Figure 3 illustrates that, upon aging, M_w- and D-related indicators increase to varying degrees. Considering SK70 asphalt as an example, the increase in M_w after PAV aging was 25.9%, 93.0%, 84.4% and 112%, respectively, when compared to the original asphalt, with an increase in the amount of crumb rubber. The increase in dispersion D was 13.6%, 74.9%, 68.1% and 88.3%, respectively. According to the results, crumb rubber modified asphalt had a significantly higher macromolecular content after aging. This is primarily due to the deep degradation of the crumb rubber during PAV aging, which releases a significant amount of macromolecular substance into the asphalt phase. This causes a significant increase in the final M_w, and a relative decrease in the content of intermediate and small molecules, as well as a greater dispersion D. A similar trend was observed with the M_w and D of JB70 and JB90 crumb rubber modified asphalt; therefore, this will not be mentioned here.

In order to further explain the relationship between the above GPC parameters, further analysis was carried out. Figure 4 illustrates the relationship between the M_w and D of crumb rubber modified asphalt. It can be seen from Figure 4a that the M_w and D were highly correlated under different aging degrees, and the goodness-of-fit R2 reached 0.9847, indicating that the M_w and D of crumb rubber modified asphalt are directly related. Due to the fact that the M_w is more sensitive to macromolecules, the degradation of crumb rubber in asphalt leads to a large increase in the content of macromolecules in asphalt. Furthermore, this trend is more obvious with increased aging. Figure 4b shows the change trend in different crumb rubber concentrations. From this figure, it can be clearly seen that for different asphalt sources and different aging conditions, there was a relatively obvious distribution interval between M_w and D under different crumb rubber concentrations, indicating that crumb rubber plays an important role in the change in asphalt M_w. Figure 4c,d illustrates the relationship between M_n and D. From this, we can see that, compared with M_w and D, the dispersion of the two was larger. M_n is more sensitive to intermediate and small molecules, indicating that the addition of crumb rubber has more influence on the change in macromolecular content in the asphalt phase and, subsequently, the change in macroscopic properties. This result was basically consistent with previous research results [30,31].

The relationship between M_w, M_n and D and the contents of the macromolecules, intermediate molecules and small molecules of asphalt were further analyzed and explored. The connection between M_w, LMS and SMS is depicted in Figure 5; it is clear that there is a strong correlation between M_w and LMS, and that the M_w of asphalt increased as LMS content increased, further highlighting the correlation between the M_w of asphalt and LMS content. For (b), it is evident that there was a negative correlation between SMS content and the M_w in asphalt, indicating that the relative increase in LMS content in asphalt caused a corresponding drop in SMS content, thus causing a negative correlation with M_w. However, the discrete type was greater than LMS. Figure 6 depicts the connection between D, LMS and SMS. Due to the high correlation between D and M_w, the pattern between D, LMS and SMS was similar to that of M_w; therefore, it will not be mentioned here.

Since M_n represents the molecular weight of the measured asphalt sample and is more sensitive to the intermediate and small molecular weight, the relationship between M_n and LMS, and MMS and SMS is represented in Figure 7. It can be seen that there was a negative correlation between M_n and SMS, and the goodness-of-fit R2 was 0.7581, with a high correlation. With the increase in SMS content, a decrease in M_n was observed. It can also be seen from Figure 7c that with increased aging, the content of SMS in asphalt gradually decreased, indicating that the saturated fraction, the aromatic fraction and other small molecular substances in asphalt shift to larger molecular substances. The relationship between M_n, and LMS and MMS also showed a certain correlation; however, it had a lower fitting quality compared to SMS. This phenomenon suggests that M_n is mainly affected by SMS content.

Figure 8 illustrates the relationship between LMS and SMS for various asphalt samples in terms of M_z. M_z was exponentially positively associated with LMS, and M_z increased as LMS content increased, according to various aging conditions. SMS and M_z had an exponentially negative correlation, and M_z declined as SMS content increased. This demonstrates that LMS content had a major impact on the M_z value of the asphalt sample. It can be seen from Figure 8a (left) that the size of M_z had little relationship to the aging degree of asphalt, and the distribution of M_z values under the three aging conditions was relatively consistent.

3.3. GPC Analysis of Filtered Asphalt

In order to analyze the influence of crumb rubber on the molecular weight of asphalt, the molecular weight and relevant information relating to asphalt obtained after the crumb rubber was filtered out was obtained. Figure 9 highlights the molecular weight distribution of three kinds of base asphalt, including 20% crumb rubber filtered asphalt and 20% crumb rubber modified asphalt after aging. It can be seen that the molecular weight distribution range of the base asphalt was smaller than that of the filtered asphalt and crumb rubber modified asphalt. The molecular weight distribution range of the equivalent asphalt after PAV aging was noticeably wider, suggesting that the molecular weight increased by various degrees. This demonstrates how crumb rubber directly affects the molecular weight of asphalt, in terms of both change and distribution. M_w, D, LMS, MSS and SMS data were further evaluated in order to quantitatively analyze the impact of crumb rubber on asphalt’s molecular weight.

According to the formula, the crumb rubber effect (CRE) and its distribution in IE and PE were calculated. Figure 10a shows the distribution of M_w in the three original asphalt states. It can be observed that the M_w of the base asphalt was the smallest, followed by crumb rubber filtered asphalt (the asphalt obtained after filtering the crumb rubber); the M_w of the corresponding crumb rubber modified asphalt was the largest. At the same time, with the increase in the amount of crumb rubber, the M_w also increased by varying degrees. Figure 10b shows that an increase in CR content had a favorable effect on IE, mostly as a result of the increase in CR content, meaning that more crumb rubber components blended into the asphalt phase. The fact that IE was significantly larger than PE for SK70 asphalt suggests that the degradation of crumb rubber into the asphalt phase had the greatest impact on M_w. The slight difference between IE and PE for JB70 and JB90 asphalt may have been due to the influence related to the type of asphalt.

Figure 11 shows the CRE of M_w under PAV and its distribution in IE and PE. Similar to the original asphalt, the M_w of the base asphalt was the smallest, followed by the crumb rubber filtered asphalt; the M_w of the corresponding crumb rubber modified asphalt was the largest. At the same time, with the increase in crumb rubber content, the M_w also increases by varying degrees. It can be seen from Figure 11b that the increase in CR content had a positive impact on IE, which was mainly due to the increase in the CR content causing more crumb rubber substances to be incorporated into the asphalt phase. For SK70 asphalt, IE was much larger than PE, indicating that M_w was mainly affected by the degradation of crumb rubber into the asphalt phase. For JB70 and JB90 asphalts, the gap between IE and PE was small. The different conditions affecting the M_w of the three types of asphalt mentioned above demonstrates that asphalt type has an important impact on the interaction between IE and PE.

4. Conclusions

In conclusion, analysis of GPC data for crumb rubber modified asphalt and extracted asphalt led to the following conclusions:

(1)

After aging of crumb rubber modified asphalt, the macromolecular content increases significantly. After PAV aging, the crumb rubber itself undergoes deep degradation, and a large number of macromolecular substances are incorporated into the asphalt phase. This results in an increase in M_w, a relative decrease in the content of intermediate and small molecules, and a greater dispersion D.

(2)

The M_w and D of crumb rubber modified asphalt are directly related. Since M_w is more sensitive to macromolecules, and M_n is more sensitive to intermediate and small molecules, this shows that the addition of crumb rubber has more influence on the change in asphalt macromolecule content. This therefore affects the change in asphalt macroscopic low-temperature performance.

(3)

The M_w of crumb rubber modified asphalt is positively correlated with LMS content, while SMS content is negatively correlated with M_w.

(4)

With the increase in crumb rubber content, M_w significantly increases, indicating that the molecular weight distribution of crumb rubber modified asphalt is affected by the reaction degree of crumb rubber in asphalt. With the increased aging degree and the increase in crumb rubber content, LMS molecules increase by varying degrees. This indicates that crumb rubber undergoes a complex physical and chemical reaction in asphalt, and consequently has a direct impact on the external macro low-temperature performance of asphalt.

(5)

The increase in CR content has a positive impact on IE, and occurs mainly due to the increase in CR content causing more crumb rubber substances to blend into the asphalt phase. For SK70 asphalt, IE is much larger than PE, indicating that M_w is mainly affected by the degradation of crumb rubber into the asphalt phase. For JB70 and JB90 asphalt, the gap between IE and PE is small, which may be caused by the influence of asphalt type.