Digital Commons @ Michigan Tech Digital Commons @ Michigan Tech Laboratory evaluation of the residue of rubber-modified Laboratory evaluation of the residue of rubber-modified emulsified asphalt emulsified asphalt

: Emulsiﬁed asphalt has been widely used in various surface treatment methods such as chip seal for low-volume road preservation. Using modiﬁed emulsiﬁed asphalt made it possible to use chip seal technology on medium- and even high-volume tra ﬃ c pavements. The main objective of the study is to quantify the residue characteristics of rubber-modiﬁed emulsiﬁed asphalt and to assess the e ﬀ ectiveness of using crumb rubber to modify emulsiﬁed asphalt binder. The four emulsiﬁed asphalt residues used the distillation procedure. Then, the rheology characteristics of emulsiﬁed asphalt residue were evaluated. The Fourier transform infrared spectroscopy (FTIR) test analyzed the chemical change of emulsiﬁed asphalt during the aging procedure. The results indicate that the evaporation method cannot remove all the water in emulsiﬁed asphalt. The mass change during the rolling thin ﬁlm oven (RTFO) process only represented the component change of emulsiﬁed asphalt binder residue. Both the high-temperature and low-temperature performance grade of the two emulsiﬁed asphalt binders with rubber were lower. The original asphalt binder adopted to emulsiﬁcation had a crucial inﬂuence on the performance of emulsiﬁed asphalt. The rubber modiﬁcation enhanced the property of the emulsiﬁed asphalt binder at low temperatures, and the improvement e ﬀ ect was enhanced as the rubber content in the emulsiﬁed asphalt was raised. The C = O band was more e ﬀ ective in quantifying the aging condition of the residue. The ﬁndings of this study may further advance the emulsiﬁed asphalt technology in pavement construction and maintenance.


Introduction
Emulsified asphalt is produced by dispersing asphalt globules in water with surfactant and some other minor ingredients [1,2]. Emulsified asphalt is mainly used for low-volume pavement preservation, which includes chip seal, tack coat, and hot-in-place recycling [3]. Currently, more than 10% of asphalt in asphalt pavement is used in the form of emulsified asphalt [4]. Emulsified asphalt can effectively decrease asphalt viscosity at low temperatures. The energy depletion and carbon emissions during the production process can be significantly reduced when compared with the conventional hot mix asphalt mixture [5].
The application of emulsified asphalt in the chip seal was prominent for the preservation maintenance of low-volume pavement. The improvement of emulsion formulation made it possible to use chip seal technology on medium and high traffic conditions [6]. Modified asphalt emulsion with styrene-butadiene rubber and natural rubber could remarkably improve the stiffness of the emulsion residue and decrease the temperature sensitivity [7][8][9][10]. Tire rubber-modified asphalt emulsion had better aging resistance than the conventional and polymer-modified asphalt emulsion. The rheological and moisture stability of tire rubber-modified asphalt emulsion had comparable or and medium temperatures had the potential of water not being totally removed. The Karl Fischer (KF) titration method could be used to quantify the water in the emulsified asphalt residue [33]. In terms of the ASTM D6934, a certain weight of emulsified asphalt was placed in the baker and heated in the oven for 3 h at 163 • C [34]. In line with the ASTM D6997, emulsified asphalt was heated in an aluminum alloy to 260 • C for 15 min to acquire the emulsified asphalt residue [35]. The high temperature guaranteed that the water can be totally removed from the emulsified asphalt. Mitchell et al. adopted the hot oven procedure, stirred-can procedure, and warm oven method to obtain the residue. Size exclusion chromatography was applied to evaluate the existence of water, proving that there was no water residue in the recovered asphalt binder [36]. Salomon used distillation, evaporation, and moisture analyzer balance procedures to obtain the residue. The rheological property and the influence of aging during the recovery procedure of three residues were different [37]. The evaporation method also affected the rheological properties of residues [38]. Islam et al. assessed the influence of curing times, temperatures, and pressures on the property of the emulsified asphalt residue [39]. There is no agreement on the most effective recovery method or which method could represent the real asphalt residue in the field. Different methods may be needed for different types of modified emulsified asphalt. This paper will follow the ASTM D6934 and the ASTM D6997 to recover emulsified asphalt and assess the emulsified asphalt residue.

Motivations and Objectives
The paper is motivated by the application of ground tire rubber to modify emulsified asphalt. Using rubber-modified emulsified asphalt could relieve the influence of waste tire on the environment, which provided a cost-effective strategy to increase the adoption of rubber on the pavement.
The main objective of the study was to quantify the residue characteristics of rubber-modified emulsified asphalt and to assess the effectiveness of using crumb rubber to modify emulsified asphalt binder, thus extending the application of crumb rubber in the asphalt pavement. The property of emulsified asphalt with rubber was compared with that of the emulsified asphalt with polymer and conventional emulsified asphalt. The four emulsified asphalts used in wet-freeze climates were analyzed. The efficiency of the distillation and evaporation test procedure was assessed. The rheological property and the low-temperature property of emulsified asphalt residue were analyzed. The effect of aging on the emulsified asphalt residue was also discussed.

Materials
Four emulsified asphalts-CRS-2P, CSS-1H, CRS-2M, and CRS-2TR-were used in the study. CRS-2P is polymer-modified emulsified asphalt. CRS-2M is styrene-butadiene-styrene (SBS)-modified emulsified asphalt with 2% rubber. CRS-2TR is tire rubber-modified emulsified asphalt. The tire rubber content in CRS-2TR emulsified asphalt is 5%. CSS-1H is conventional emulsified asphalt. The technical performance of the four emulsified asphalt binders is displayed in Table 1. The test materials and test methods are displayed in Table 2.

Test Methods
The distillation and evaporation methods were adopted to obtain CSS-1H residue. The efficiency of the two methods in removing water in emulsified asphalt was assessed. The more effective method, the distillation method, was used to obtain the residue of different emulsified asphalt binders. The performance of emulsified asphalt residue was assessed using different test methods. In order to guarantee the repeatability of the test results, we used at least three samples for each test.
a. Emulsified asphalt distillation Emulsified asphalt binders were heated in an aluminum alloy still to 260 • C to obtain the emulsified asphalt residue by following ASTM D6997. The distilled emulsified asphalt residue was collected for further testing.
b. Residue obtain methods comparison In order to compare the efficiency of the evaporation and distillation procedure in removing the water in emulsified asphalt residue, we conducted two procedures on the CSS-1H. The evaporation residue was collected by heating CSS-1H at 163 • C for 3 h to expel water in the emulsified asphalt, which followed the ASTM D6934 procedure.
c. Rolling thin film oven (RTFO) The RTFO test was adopted to measure the influence of air and heat on the aging property of emulsified asphalt residue, according to the ASTM D2872. The moving film of emulsified asphalt residue was heated in an oven for 85 min at 163 • C [40]. The mass change of emulsified asphalt residue can be determined, which can reflect the volatility and the oxidation of the residue. d. Pressure aging vessel (PAV) The PAV test represented the accelerated aging of emulsified asphalt residue under compressed air and elevated temperature, which is consistent with the ASTM D6521. The RTFO-aged emulsified asphalt residue was aged at 100 • C for 20 h with 2.10 MPa air pressure [41]. e. Dynamic shear rheometer (DSR) The DSR test was used to quantify the residue rheological property of emulsified asphalt. The complex shear modulus (|G*|) and phase angle (δ) of the sample are acquired with the DSR test [42]. The DSR temperature sweeps of unaged and RTFO-aged residues were conducted from 34 • C to 82 • C with a 6 • C increment. For the PAV-aged asphalt binder, we conducted the DSR temperature sweeps from 25 • C to 13 • C with a 3 • C decrement. Six frequencies (0.1, 1, 1.59, 3, 5, and 10 Hz) were adopted for every temperature sweep of residues experiencing the different aging procedure. The test parameters of the DSR test are displayed in Table 1. f. Asphalt binder cracking device (ABCD) The cracking temperature of emulsified asphalt residue was obtained by the ABCD test, and the test was followed by the AASHTO TP92. The fracture stress of the residue when cracking could directly reflect the low-temperature property of the residue [43]. The ABCD test can also assess the low-temperature property improvement because of modification [44,45]. The ABCD test was conducted on the ABCD test apparatus from EZ Asphalt Technology. The cracking temperature of emulsified asphalt residue was determined when strain jump occurred during the cooling process [46]. The fracture g. FTIR The chemical interaction of emulsified asphalt residue could be validated with the FTIR test method [47]. The functional group could be adopted as an indicator to assess the asphalt modification effect [48,49]. The Jasco FT/IR 4200 FTIR spectrometer was used to quantify the chemical components of emulsified asphalt residue. The spectra wavelength ranged from 600 to 4000 cm −1 with a resolution of 4 cm −1 . For every sample, we used 64 scan times to guarantee the stability of the results, as displayed in Table 1. The -OH (3200-3550 cm −1 ) functional group was selected for the CSS-1H residue with evaporation and distillation method to compare the efficiency of the two methods in removing the water in emulsified asphalt residue. The changing of C=O and S=O groups could be used to quantify the aging effect [50]. Bands near 1700 cm −1 were the stretch of C=O, and bands near 1000 cm −1 were the movement of S=O [51].

The Comparison of Evaporation and Distillation Procedure
The -OH (3200-3550 cm −1 ) functional group was selected to validate the existence of water in residue obtained with the distillation and evaporation methods. The FTIR test results of two CSS-1H residues are shown in Figure 1. The CSS-1H residue after the evaporation procedure peaked during 3200-3550 cm −1 , which proved the existence of water in the residue. However, there was no peak during 3200-3550 cm −1 for CSS-1H after the distillation procedure. Using the evaporation method to obtain the emulsified asphalt residue cannot remove all the water in emulsified asphalt, thus influencing the property of emulsified asphalt residue. The characteristics evaluation of the emulsified asphalt of the following tests was conducted on residue acquired with the distillation method. The chemical interaction of emulsified asphalt residue could be validated with the FTIR test method [47]. The functional group could be adopted as an indicator to assess the asphalt modification effect [48,49]. The Jasco FT/IR 4200 FTIR spectrometer was used to quantify the chemical components of emulsified asphalt residue. The spectra wavelength ranged from 600 to 4000 cm −1 with a resolution of 4 cm −1 . For every sample, we used 64 scan times to guarantee the stability of the results, as displayed in Table 1. The -OH (3200-3550 cm −1 ) functional group was selected for the CSS-1H residue with evaporation and distillation method to compare the efficiency of the two methods in removing the water in emulsified asphalt residue. The changing of C=O and S=O groups could be used to quantify the aging effect [50]. Bands near 1700 cm −1 were the stretch of C=O, and bands near 1000 cm −1 were the movement of S=O [51].

The Comparison of Evaporation and Distillation Procedure
The -OH (3200-3550 cm −1 ) functional group was selected to validate the existence of water in residue obtained with the distillation and evaporation methods. The FTIR test results of two CSS-1H residues are shown in Figure 1. The CSS-1H residue after the evaporation procedure peaked during 3200-3550 cm −1 , which proved the existence of water in the residue. However, there was no peak during 3200-3550 cm −1 for CSS-1H after the distillation procedure. Using the evaporation method to obtain the emulsified asphalt residue cannot remove all the water in emulsified asphalt, thus influencing the property of emulsified asphalt residue. The characteristics evaluation of the emulsified asphalt of the following tests was conducted on residue acquired with the distillation method.

The Mass Change during the RTFO Test
The evaporation of light compositions in residue reduced the mass, and the mass raised when it interacted with oxygen. The mass change during the RTFO process is due to the synthetic interaction of the volatilization and oxidation process. The mass change of four emulsified asphalt residues after the RTFO test is displayed in Figure 2. The standard deviation (SD) of the mass loss for the test results are also included in Figure 2. On the basis of the ASTM D6373, the mass change should be less than 1% during the RTFO aging procedure for asphalt binders. The mass of four asphalt residues

The Mass Change during the RTFO Test
The evaporation of light compositions in residue reduced the mass, and the mass raised when it interacted with oxygen. The mass change during the RTFO process is due to the synthetic interaction of the volatilization and oxidation process. The mass change of four emulsified asphalt residues after the RTFO test is displayed in Figure 2. The standard deviation (SD) of the mass loss for the test results are also included in Figure 2. On the basis of the ASTM D6373, the mass change should be less than 1% during the RTFO aging procedure for asphalt binders. The mass of four asphalt residues decreased during the short-term aging procedure. The mass of CSS-1H and CRS-2TR residues decreased by more than 1%, which may have been because of the evaporation of surfactant in the emulsified asphalt. The FTIR test of four unaged residues during 3200-3550 cm −1 is displayed in Figure 3. No peak occurred during 3200-3550 cm −1 in the four emulsified asphalt residues acquired using the distillation process. The mass change of the emulsified asphalt binder during the RTFO process was not caused by the evaporation of water. The mass change only represented the components change of emulsified asphalt binder residue.  The FTIR test of four unaged residues during 3200-3550 cm −1 is displayed in Figure 3. No peak occurred during 3200-3550 cm −1 in the four emulsified asphalt residues acquired using the distillation process. The mass change of the emulsified asphalt binder during the RTFO process was not caused by the evaporation of water. The mass change only represented the components change of emulsified asphalt binder residue. The FTIR test of four unaged residues during 3200-3550 cm −1 is displayed in Figure 3. No peak occurred during 3200-3550 cm −1 in the four emulsified asphalt residues acquired using the distillation process. The mass change of the emulsified asphalt binder during the RTFO process was not caused by the evaporation of water. The mass change only represented the components change of emulsified asphalt binder residue.  Six different frequency sweeps under different test temperatures were applied to four types of emulsified asphalt residues. The master curve of |G*| was generated with a nonlinear regression method [52].
The master curve of the |G*| of four unaged emulsified asphalt residues is displayed in Figure 4. The property of asphalt binder at high reduced frequency could reflect the low-temperature properties. The modulus of the residue was enhanced when the temperature decreased, thus increasing the |G*|. The |G*| of CRS-2TR, CRS-2M, CRS-2P, and CSS-1H increased gradually. The |G*| differences between the CRS-2TR, CRS-2M, and CRS-2P were minimal when the reduced frequency was high. The maximum |G*| of CSS-1H was about five times higher than the maximum |G*| of CRS-2TR.  The master curve of the |G*| of four RTFO-aged emulsified asphalt residues are displayed in Figure 5. The |G*| of CRS-2TR, CRS-2M, and CRS-2P almost had no differences when the reduced frequency was high. The |G*| of CSS-1H was the highest among all residues. The |G*| of CSS-1H was two times larger at the low reduced frequency conditions and four times larger at the high reduced frequency conditions when compared with the other three emulsified asphalts. By comparing the |G*| of the test results in Figures 4 and 5, we found that the |G*| was increased after RTFO aging. The aging procedure oxidized the light component of the asphalt and increased the |G*|. The master curve of the |G*| of four RTFO-aged emulsified asphalt residues are displayed in Figure 5. The |G*| of CRS-2TR, CRS-2M, and CRS-2P almost had no differences when the reduced frequency was high. The |G*| of CSS-1H was the highest among all residues. The |G*| of CSS-1H was two times larger at the low reduced frequency conditions and four times larger at the high reduced frequency conditions when compared with the other three emulsified asphalts. By comparing the |G*| of the test results in Figures 4 and 5, we found that the |G*| was increased after RTFO aging. The aging procedure oxidized the light component of the asphalt and increased the |G*|.
The master curve of the |G*| of the four PAV-aged emulsified asphalt residues are displayed in Figure 6. The |G*| of CRS-2P, CRS-2TR, CRS-2M, and CSS-1H increased gradually. The differences between the |G*| of different emulsified asphalt residues was increased after the PAV aging. The |G*| of CRS-2TR was similar to the |G*| of CRS-2M when the reduced frequency was low, and the |G*| of CRS-2P was similar to the |G*| of CRS-2TR when the reduced frequency was high. This may be because of the existence of 5% tire rubber in CRS-2TR. The |G*| of CSS-1H was four times larger than the |G*| of CRS-2P. frequency was high. The |G*| of CSS-1H was the highest among all residues. The |G*| of CSS-1H was two times larger at the low reduced frequency conditions and four times larger at the high reduced frequency conditions when compared with the other three emulsified asphalts. By comparing the |G*| of the test results in Figures 4 and 5, we found that the |G*| was increased after RTFO aging. The aging procedure oxidized the light component of the asphalt and increased the |G*|. The master curve of the |G*| of the four PAV-aged emulsified asphalt residues are displayed in Figure 6. The |G*| of CRS-2P, CRS-2TR, CRS-2M, and CSS-1H increased gradually. The differences between the |G*| of different emulsified asphalt residues was increased after the PAV aging. The |G*| of CRS-2TR was similar to the |G*| of CRS-2M when the reduced frequency was low, and the |G*| of CRS-2P was similar to the |G*| of CRS-2TR when the reduced frequency was high. This may be because of the existence of 5% tire rubber in CRS-2TR. The |G*| of CSS-1H was four times larger than the |G*| of CRS-2P. The (|G*|/sinδ) of unaged emulsified asphalt under 10 rad/s angular frequency are displayed in Figure 7. The (|G*|/sinδ) of residues reduced when the test temperature was raised. The rutting parameter of the CSS-1H was remarkably higher than that of other emulsified asphalts. The CSS-1H passed the standard restriction (higher than 1.0 kPa) at 70 °C. The CRS-2P met the standard at 64 °C. The two emulsified asphalts, CRS-2M and CRS-2TR, fulfilled the criteria at 58 °C.

The Rutting Parameter (|G*|/sinδ) and the Fatigue Parameter (|G*|·sinδ)
The (|G*|/sinδ) of unaged emulsified asphalt under 10 rad/s angular frequency are displayed in Figure 7. The (|G*|/sinδ) of residues reduced when the test temperature was raised. The rutting parameter of the CSS-1H was remarkably higher than that of other emulsified asphalts. The CSS-1H passed the standard restriction (higher than 1.0 kPa) at 70 • C. The CRS-2P met the standard at 64 • C. The two emulsified asphalts, CRS-2M and CRS-2TR, fulfilled the criteria at 58 • C.
The rutting parameter of RTFO-aged emulsified asphalts under 10 rad/s angular frequency is displayed in Figure 8. The rutting parameter of the RTFO-aged binder had a trend similar to the unaged binder. The critical temperatures for the RTFO-aged samples of CSS-1H, CRS-2P, and CRS-2TR were the same as the unaged samples. The rutting parameter of RTFO-aged CRS-2M was higher than 2.2 kPa, even at 64 • C. The CRS-2M had better rutting resistance than CRS-2TR under the RTFO aging condition, which may have been because of the lower rubber content in CRS-2M. The (|G*|/sinδ) of unaged emulsified asphalt under 10 rad/s angular frequency are displayed in Figure 7. The (|G*|/sinδ) of residues reduced when the test temperature was raised. The rutting parameter of the CSS-1H was remarkably higher than that of other emulsified asphalts. The CSS-1H passed the standard restriction (higher than 1.0 kPa) at 70 °C. The CRS-2P met the standard at 64 °C. The two emulsified asphalts, CRS-2M and CRS-2TR, fulfilled the criteria at 58 °C.  The rutting parameter of RTFO-aged emulsified asphalts under 10 rad/s angular frequency is displayed in Figure 8. The rutting parameter of the RTFO-aged binder had a trend similar to the unaged binder. The critical temperatures for the RTFO-aged samples of CSS-1H, CRS-2P, and CRS-2TR were the same as the unaged samples. The rutting parameter of RTFO-aged CRS-2M was higher than 2.2 kPa, even at 64 °C. The CRS-2M had better rutting resistance than CRS-2TR under the RTFO aging condition, which may have been because of the lower rubber content in CRS-2M. On the basis of the DSR test, we found that the high-temperature performance grade of the CSS-1H (70 °C) was the highest, followed by CRS-2P (64 °C). The high-temperature properties of CRS-2M and CRS-2TR were the lowest (58 °C). The CSS-1H is formulated with harder base asphalt (low penetration), thus having better high-temperature properties. The original asphalt binder used for asphalt emulsification should be considered in quantifying the property of emulsified asphalt binders.
The |G*|·sinδ of PAV-aged emulsified asphalts are presented in Figure 9. CRS-2P and CRS-2TR still met the standard criteria (|G*|·sinδ lower than 5000 kPa) at 13 °C, which corresponded to a lowtemperature performance grade lower than −40 °C. For CSS-1H, the |G*|·sinδ met the criteria at the temperature of 25 °C, which corresponded to a low-temperature performance grade of −28 °C. The CRS-2M met the requirements at a temperature of 16 °C, which corresponded to the low-temperature performance grade of −34 °C. Limited results could be observed on the low-temperature property of emulsified asphalt residues using the DSR test. The ABCD test was adopted to evaluate the lowtemperature characteristics of the residues. On the basis of the DSR test, we found that the high-temperature performance grade of the CSS-1H (70 • C) was the highest, followed by CRS-2P (64 • C). The high-temperature properties of CRS-2M and CRS-2TR were the lowest (58 • C). The CSS-1H is formulated with harder base asphalt (low penetration), thus having better high-temperature properties. The original asphalt binder used for asphalt emulsification should be considered in quantifying the property of emulsified asphalt binders.
The |G*|·sinδ of PAV-aged emulsified asphalts are presented in Figure 9. CRS-2P and CRS-2TR still met the standard criteria (|G*|·sinδ lower than 5000 kPa) at 13 • C, which corresponded to a low-temperature performance grade lower than −40 • C. For CSS-1H, the |G*|·sinδ met the criteria at the temperature of 25 • C, which corresponded to a low-temperature performance grade of −28 • C. The CRS-2M met the requirements at a temperature of 16 • C, which corresponded to the low-temperature performance grade of −34 • C. Limited results could be observed on the low-temperature property of emulsified asphalt residues using the DSR test. The ABCD test was adopted to evaluate the low-temperature characteristics of the residues.

ABCD Test
The low-temperature properties of residues were assessed with the ABCD test. The thermal cracking temperature can be determined directly on the basis of the test. The fracture stress and the PG grade by ABCD can be calculated on the basis of Equations (1) and (2). The ABCD test results are displayed in Figure 10. According to the ABCD test, CRS-2M and CRS-2TR had higher strain jump and fracture stress than CRS-2P and CSS-1H. The crack temperature of the CRS-2M and CRS-2TR was lower, and the asphalt was more brittle at a lower temperature, thus having higher fracture stress. The temperature had a remarkable effect on the stress in the asphalt binder at low temperatures. The CRS-2P had a lower crack temperature than the CSS-1H, but the fracture stress of the CRS-2P was lower. Polymer modification could improve the stress relaxation ability of asphalt at low temperatures. On the basis of the ABCD test, we found that the CRS-2M and CRS-2TR emulsified asphalt binder residues had better low-temperature properties than other emulsified residues. The low-temperature characteristics of the emulsified asphalt were improved after adding rubber, and the improvement effect enhanced as the rubber content increased.

ABCD Test
The low-temperature properties of residues were assessed with the ABCD test. The thermal cracking temperature can be determined directly on the basis of the test. The fracture stress and the PG grade by ABCD can be calculated on the basis of Equations (1) and (2). The ABCD test results are displayed in Figure 10. According to the ABCD test, CRS-2M and CRS-2TR had higher strain jump and fracture stress than CRS-2P and CSS-1H. The crack temperature of the CRS-2M and CRS-2TR was lower, and the asphalt was more brittle at a lower temperature, thus having higher fracture stress. The temperature had a remarkable effect on the stress in the asphalt binder at low temperatures. The CRS-2P had a lower crack temperature than the CSS-1H, but the fracture stress of the CRS-2P was lower. Polymer modification could improve the stress relaxation ability of asphalt at low temperatures. On the basis of the ABCD test, we found that the CRS-2M and CRS-2TR emulsified asphalt binder residues had better low-temperature properties than other emulsified residues. The low-temperature characteristics of the emulsified asphalt were improved after adding rubber, and the improvement effect enhanced as the rubber content increased.

ABCD Test
The low-temperature properties of residues were assessed with the ABCD test. The thermal cracking temperature can be determined directly on the basis of the test. The fracture stress and the PG grade by ABCD can be calculated on the basis of Equations (1) and (2). The ABCD test results are displayed in Figure 10. According to the ABCD test, CRS-2M and CRS-2TR had higher strain jump and fracture stress than CRS-2P and CSS-1H. The crack temperature of the CRS-2M and CRS-2TR was lower, and the asphalt was more brittle at a lower temperature, thus having higher fracture stress. The temperature had a remarkable effect on the stress in the asphalt binder at low temperatures. The CRS-2P had a lower crack temperature than the CSS-1H, but the fracture stress of the CRS-2P was lower. Polymer modification could improve the stress relaxation ability of asphalt at low temperatures. On the basis of the ABCD test, we found that the CRS-2M and CRS-2TR emulsified asphalt binder residues had better low-temperature properties than other emulsified residues. The low-temperature characteristics of the emulsified asphalt were improved after adding rubber, and the improvement effect enhanced as the rubber content increased.  The performance-grade temperature for the four different emulsified asphalts is displayed in Table 3. CSS-1H had the best high-temperature property but had the worst low-temperature property. For CRS-2P, the high-temperature property was weaker and the low-temperature property was improved compared with CSS-1H. The high-temperature performance of the CRS-2M and CRS-2TR had the same high-temperature grade, but the low-temperature property of the CRS-2TR was better. Rubber addition could enhance the low-temperature characteristics of emulsified asphalt residue. The test samples are based on limited test samples, and more samples are needed in order to verify the conclusion. Furthermore, the influence of the original asphalt binder used to produce the emulsified asphalt binder should also be considered.

Fourier Transform Infrared (FTIR) Test
The change of chemical bonds in residues under different aging conditions were quantified with the FTIR test. The FTIR test results of different emulsified asphalt residues under different aging conditions is displayed in Figure 11. The band near 1000 cm −1 (S=O) and 1700 cm −1 (C=O) were enlarged to check the differences between different aging conditions. For CSS-1H, the S=O band had no differences between the control and short-term aged binders, and the area of S=O band after PAV aging increased dramatically. The C=O band area increased slightly after short-term aging and improved significantly after long-term aging (Figure 11a). For CRS-2P, the S=O band and the C=O band increased after RTFO and PAV aging (Figure 11b). For CRS-2M, the S=O band of the unaged binder was slightly higher than RTFO-aged binder, and the area of S=O band after PAV aging increased dramatically. The C=O band area increased slightly after short-term aging and improved significantly after long-term aging (Figure 11c). For CRS-2TR, the S=O band increased dramatically after PAV aging. The C=O band area increased slightly after short-term aging and improved significantly after long-term aging (Figure 11d).
In order to quantitatively evaluate the influence of aging on the emulsified asphalt residue, we calculated the band area of C=O and S=O under different aging conditions on the basis of the upper limit and lower limit and the baseline [53]. The spectrum area calculation results for the C=O band are shown in Figure 12. The C=O band area increased with the aging degree for all emulsified asphalt binder residues. The C=O band area of CRS-2P was the highest in four emulsified asphalt binders. The C=O band area of CRS-2TR was the lowest in four emulsified asphalt binders. The area increases for CSS-1H was lowest after the PAV aging. According to the DSR test result, the |G*| of the CSS-1H was the highest, and the stiffness of the CSS-1H was higher. The influence of long-term aging was the least.
The spectrum area calculation results for the S=O band are presented in Figure 13. The S=O band area was lower than the C=O band area. For CRS-2P, the S=O band area of RTFO-aged samples was higher than that of unaged and PAV aged samples. For CRS-2TR, the S=O band area of RTFO-aged samples was lower than that of unaged and long-term aged samples. The differences between the unaged and short-term aged asphalts were minimal for the CSS-1H and the CRS-2M. The S=O band area did not match to the aging condition of CRS-2P residue.
binder was slightly higher than RTFO-aged binder, and the area of S=O band after PAV aging increased dramatically. The C=O band area increased slightly after short-term aging and improved significantly after long-term aging (Figure 11c). For CRS-2TR, the S=O band increased dramatically after PAV aging. The C=O band area increased slightly after short-term aging and improved significantly after long-term aging (Figure 11d). In order to quantitatively evaluate the influence of aging on the emulsified asphalt residue, we calculated the band area of C=O and S=O under different aging conditions on the basis of the upper limit and lower limit and the baseline [53]. The spectrum area calculation results for the C=O band are shown in Figure 12. The C=O band area increased with the aging degree for all emulsified asphalt binder residues. The C=O band area of CRS-2P was the highest in four emulsified asphalt binders. The C=O band area of CRS-2TR was the lowest in four emulsified asphalt binders. The area increases for CSS-1H was lowest after the PAV aging. According to the DSR test result, the |G*| of the CSS-1H was the highest, and the stiffness of the CSS-1H was higher. The influence of long-term aging was the least. In order to quantitatively evaluate the influence of aging on the emulsified asphalt residue, we calculated the band area of C=O and S=O under different aging conditions on the basis of the upper limit and lower limit and the baseline [53]. The spectrum area calculation results for the C=O band are shown in Figure 12. The C=O band area increased with the aging degree for all emulsified asphalt binder residues. The C=O band area of CRS-2P was the highest in four emulsified asphalt binders. The C=O band area of CRS-2TR was the lowest in four emulsified asphalt binders. The area increases for CSS-1H was lowest after the PAV aging. According to the DSR test result, the |G*| of the CSS-1H was the highest, and the stiffness of the CSS-1H was higher. The influence of long-term aging was the least. The spectrum area calculation results for the S=O band are presented in Figure 13. The S=O band area was lower than the C=O band area. For CRS-2P, the S=O band area of RTFO-aged samples was higher than that of unaged and PAV aged samples. For CRS-2TR, the S=O band area of RTFO-aged samples was lower than that of unaged and long-term aged samples. The differences between the unaged and short-term aged asphalts were minimal for the CSS-1H and the CRS-2M. The S=O band area did not match to the aging condition of CRS-2P residue.

Conclusions
In this study, we evaluated four emulsified asphalt binders applied in wet-freeze regions. The property of rubber-modified emulsified asphalt was compared with that of polymer-modified asphalt and conventional emulsified asphalt. The distillation residue was RTFO-aged and PAV-aged in the lab. The DSR, ABCD, and FTIR tests were adopted to investigate the property of emulsified asphalt residue.
The main conclusions are presented below, on the basis of the test results: 1. The evaporation method was less effective than the distillation method in removing water from residue. 2. The mass change during the RTFO process only represented the component change of emulsified asphalt binder residue. 3. The high-temperature property of the CSS-1H was the best, and the high-temperature properties of the two emulsified asphalts with rubber were the worst. The original asphalt binder used to produce emulsified asphalt had a remarkable influence on the characteristics of emulsified asphalt. 4. Polymer modification could enhance the low-temperature property of the emulsified asphalt.
The low-temperature properties of CRS-2M and CRS-2TR residue were improved. The rubber modification enhanced the low-temperature characteristics of the emulsified asphalt binder, and the improvement effect enhanced as the rubber content increased. 5. The C=O band was more sensitive to the aging condition of the residue than S=O, which can be used as an index to quantify the aging condition of the residue.

Conflicts of Interest:
The authors declare no conflict of interest.

Conclusions
In this study, we evaluated four emulsified asphalt binders applied in wet-freeze regions. The property of rubber-modified emulsified asphalt was compared with that of polymer-modified asphalt and conventional emulsified asphalt. The distillation residue was RTFO-aged and PAV-aged in the lab. The DSR, ABCD, and FTIR tests were adopted to investigate the property of emulsified asphalt residue.
The main conclusions are presented below, on the basis of the test results: 1.
The evaporation method was less effective than the distillation method in removing water from residue.

2.
The mass change during the RTFO process only represented the component change of emulsified asphalt binder residue.

3.
The high-temperature property of the CSS-1H was the best, and the high-temperature properties of the two emulsified asphalts with rubber were the worst. The original asphalt binder used to produce emulsified asphalt had a remarkable influence on the characteristics of emulsified asphalt.

4.
Polymer modification could enhance the low-temperature property of the emulsified asphalt. The low-temperature properties of CRS-2M and CRS-2TR residue were improved. The rubber modification enhanced the low-temperature characteristics of the emulsified asphalt binder, and the improvement effect enhanced as the rubber content increased. 5.
The C=O band was more sensitive to the aging condition of the residue than S=O, which can be used as an index to quantify the aging condition of the residue. Funding: This study is sponsored and partially funded by the Michigan Department of Environment, Great Lakes, and Energy (EGLE) of the United States.