Construction and Building Materials 101 (2015) 536–547

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Chemo-physical analysis and molecular dynamics (MD) simulation of

moisture susceptibility of nano hydrated lime modified asphalt mixtures
Hui Yao, Qingli Dai, Zhanping You ⇑
Department of Civil and Environmental Engineering, Michigan Technological University, 1400 Townsend Drive, Houghton, MI 49931, USA

h i g h l i g h t s

Carboxylic acids, anhydrides, and ketones are the major products in the asphalt.

The aging of asphalt improves the resistance to moisture damage.

Interface MD models were generated to simulate the mechanism of moisture damage.

a r t i c l e i n f o a b s t r a c t

Article history: The purpose of this study is to investigate the moisture susceptibility of nano hydrated lime (NHL)
Received 6 May 2015 modified asphalt mixtures and determine the fundamental factors of moisture damage in asphalt mix-
Received in revised form 28 August 2015 tures using the molecular dynamics (MD), as well as analyze the effect that the aging of asphalt binders
Accepted 15 October 2015
has on the moisture damage in asphalt mixtures. The NHL was added to the control asphalt to make NHL
modified asphalt. The modified asphalt was mixed with aggregates to form NHL modified asphalt
mixtures. The tensile strength ratio (TSR) test was used to evaluate the moisture susceptibility of asphalt
Keywords:
mixtures. When the TSR test was done, different solutions were used to extract the polar groups from the
Asphalt
Nano hydrated lime (NHL)
tested asphalt mixtures, and the polar groups were analyzed using Fourier Transform Infrared
Fourier Transform Infrared Spectroscopy Spectroscopy (FTIR) attenuated total reflection (ATR). When asphalt oxidizes, the six functional groups
(FTIR) (ketones, carboxylic acids, anhydrides, aldehydes, amides and esters) with the carbonyl group can be
Moisture susceptibility found in the polar part of asphalt based on FTIR ATR spectral data and references. The polar groups with
Aging effect the carbonyl group also indicate the oxidation extent of asphalt binders. The FTIR test results indicate
Molecular dynamics (MD) that carboxylic acids and ketones were the primary aging products in asphalt, and these two carbonyl
Chemo-physical analysis groups relate to the rutting resistance and moisture susceptibility of asphalt mixtures. Furthermore,
Nano modified asphalt
the MD interface systems of asphalt–aggregate and aggregate–water were created, as well as the sepa-
Modified asphalt
rated systems (asphalt, water, and aggregates). The essential mechanisms for the moisture susceptibility
of asphalt mixtures were explored with MD simulations. The potential energies of the interface systems
and each separated system were computed to obtain the adhesion energies of the interface MD models.
The differences in adhesion energy between the asphalt–aggregate and aggregate–water systems show
that water tends to bond to the aggregate rather than the asphalt. This also explains the displacement
of water that occurs in asphalt mixtures when there is moisture damage. In addition, the effect of asphalt
aging on the moisture susceptibility of the asphalt mixture was also analyzed using MD interface models.
The MD results demonstrate that the aging group in asphalt is helpful in reducing the moisture damage in
asphalt mixtures.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction heat due to nature are the main causes of asphalt aging. The car-
bonyl and sulfoxide groups increase in the asphalt as oxidation
The oxidation hardens the asphalt and increases the viscosity of occurs. The carbonyl groups include ketones, carboxylic acids,
the asphalt. The exposure to ultraviolet radiation (UV), oxygen, and anhydrides, aldehydes, amides and esters. During the aging of
asphalt, the aromatic components of the asphalt agglomerated
⇑ Corresponding author. and caused a reduction in the mobility and reactivity of the asphalt
E-mail addresses: [email protected] (H. Yao), [email protected] (Q. Dai), [1–3]. An earlier study on asphalt aging with oxygen focused on
[email protected] (Z. You).

http://dx.doi.org/10.1016/j.conbuildmat.2015.10.087
0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.
 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547 537

the relationship between the properties of asphalt and the extent 2. Research objectives
of oxidation. Asphaltene is the main component that affects the
viscosity of asphalt [4]. Oxidation increased the spatial variations The objectives of this study are to analyze the effect aging of the
of asphalt on the nanoscale. The aging of asphalt also increased NHL modified asphalt binders has on the moisture damage in
the adhesive and cohesive strengths of the asphalt in the early asphalt mixtures, and to understand the fundamental factors or
aging state. This indicates that the resistance to rutting and the chemical groups in the asphalt, which link to the moisture suscep-
moisture susceptibility in asphalt mixtures improved. However, tibility of asphalt mixtures. The functional groups with the carbonyl
the adhesive and cohesive strengths of the asphalt were not group in the asphalt can be found to relate to the water damage of
enhanced after aging with the pressure aging vessel (PAV). The asphalt mixtures. The six functional groups (ketones, carboxylic
micromechanical properties of asphalt were determined by the acids, anhydrides, aldehydes, amides and esters) containing the
asphaltene content and the size of the microstructures [5]. In carbonyl group in the asphalt were demonstrated and studied.
addition, some modifiers have an anti-aging performance, such Furthermore, the molecular dynamics were applied to analyze the
as crumb rubber with styrene–butadiene–styrene (SBS), and nano- relationships and interactions of three interface systems: asphalt–
materials [6,7]. aggregate system, aggregate–water system and aggregate–asphalt
Modified asphalt binders have been widely used in road con- with the carboxylic acid (aging group), as well as the impact of
struction to improve the performance of pavement. Researchers asphalt aging on the moisture damage in asphalt mixtures.
have been searching for innovative materials to improve the
performance of asphalt binders and asphalt mixtures. The use of 3. Materials and experimental design
non-modified nanoclay (NMN) and polymer modified nanoclay
(PMN) in the control asphalt binder improved the complex shear 3.1. Chemistry of carbonyl groups

modulus (G⁄) and viscosity [7–9]. The micro- or nano-sized

During asphalt oxidation, the six functional groups with the carbonyl group are
materials (nanoclay and carbon microfiber) improved the moisture formed including carboxylic acids, aldehydes, amides, anhydrides, esters, and
susceptibility of the modified asphalt mixtures or decreased the ketones. (1) Carboxylic acid in the organic compound consists of a carboxyl group
potential of moisture damage [10]. The recycled asphalt pavement (C (O) OH) and R (the rest of the molecules) [24], and it is easily detected and iden-
tified by the FTIR. Carboxylic acid is a weak acid and can only donate H+ cations
(RAP) was also used to reduce the moisture susceptibility of the hot
[25]. (2) Ketones are the organic components with the structure of RC (@O) R0 ,
mix asphalt (HMA) and warm mix asphalt (WMA) [11–13]. In where the R and R0 are the hydrocarbon groups. They do not have a reactive group
accordance with the literature review [14–18], hydrated lime due to their structures [25]. The ketone carbon is often demonstrated as being ‘‘sp2
improved the moisture susceptibility of asphalt mixtures. The addi- hybridized” (electron and molecular structures), and ketones are trigonal and pla-
tion of the hydrated lime allowed for the precipitation of calcium nar around the ketonic carbons. The ketones are different from the functional
groups with a carbonyl group, such as carboxylic acids, amides, and esters [26].
ions onto the aggregate surface, making it more favorable to the
(3) Anhydrides consist of two acyl groups (RCO–, carbonyl and alkyl groups) con-
asphalt binder and improving the binder-to-aggregate adhesion nected to the oxygen atom in the acid anhydrides [24]. It is also a reactive acyl
[19,20]. The hydrated lime was either added into the asphalt group, so it can produce carboxylic acid and acetate esters [27]. 4) Aldehydes are
directly or into the asphalt mixtures as filler, and both ways were also an organic group with the structure R–CHO. This structure is called the formyl
group when the R is replaced by H. The sp2 hybridized and planar carbon connects
shown to improve the resistance to moisture damage and frost, as
oxygen with a double bond. The aldehydes and carboxylic acids can be produced
well as a resistance to chemical aging [16]. That was the motivation from the oxidations of alcohol, and they are polar groups [24,28,29]. (5) Amides
to select the nano hydrated lime (NHL) in this study. are a group with the structure RnEOxNR0 2, where E represents different elements
Molecular mechanics is when classical mechanics (Newtonian and n and x denote the number of atoms [24]. Amide can be the hydrogen bond
mechanics) is used to model molecular systems. The appropriate acceptor and donor, due to the carbonyl bond. This characteristic of the hydrogen
bond acceptor and donor also makes it soluble in water [30]. (6) Esters are chemical
force field is assigned to the molecular systems, and the potential
compounds with the structure RCO2R0 , where R and R0 are the hydrocarbons in the
energies of these systems are calculated. The physical movements organic esters. Esters can be derived from the condensation of the carboxylic acids,
of atoms and molecules are simulated by the computer program in such as formate, hexyl octanoate and acetate. The reactions include alcoholysis of
the molecular dynamics (MD) simulation. These atoms and mole- acyl chlorides and acid anhydrides, esterification of carboxylic acids, and so on
[24,31–33].
cules move and interact to mimic the response to external forces
or environments. The MD program, Large-scale Atomic/Molecular
3.2. Preparation of polar groups in the asphalt
Massively Parallel Simulator (LAMMPS) [21], is used to study the
interaction between the asphalt and aggregates, as well as the The control asphalt used was a PG 58-28 asphalt binder. Nano hydrated lime
reciprocity between the water and aggregates. Five kinds of MD (NHL) with a particle size of 100 nm (Fig. 1a) was shipped from, Sha Highland Net-
simulations are usually used to model the different states in the work Tech Company of China, in Shanghai city. The micro-image of NHL was
observed by the field emission scanning electron microscopy (FE-SEM). In Fig. 1b,
systems, such as microcanonical ensemble (NVE ensemble), canon-
the NHL materials were agglomerated. The NHL consists of 90% calcium hydroxide,
ical ensemble (NVT ensemble), isothermal–isobaric ensemble (NPT 3% calcium carbonate, 1.5% calcium oxide, 1.5% silicon dioxide and some insoluble
ensemble), isoenthalpic–isobaric ensemble (NPH ensemble), and matters. The density of NHL is 2.24 g/cm3. Amounts of 10% and 20% NHL by weight
generalized ensembles [22]. The relationship between the asphalt of asphalt were added to the control asphalt.
and aggregate was developed, and the asphalt–quartz structure The NHL material was slowly added into the control asphalt and the modified
asphalt binders were blended for two hours in the high shear machine. The asphalt
model of the interface was used in the system [23]. The Consistent binders and aggregates were blended in the drum mixer and compacted by a Super-
Valence force field (CVFF_aug) was considered in this simulation to pave Gyratory Compactor (SGC). Then, the indirect tensile test of asphalt mixtures
characterize the inter-atom interactions. The shear stress and vis- was performed, and the tensile strength ratio (TSR) of asphalt mixtures was evalu-
cosity were evaluated and acceptable simulation results were ated. When the TSR test was done, the tested samples were immersed and extracted
by different solutions. Firstly, the cyclohexane solution was used to extract the non-
obtained using the CVFF_aug force field [23].
polar part of the asphalt in the tested asphalt mixtures. Then, the toluene and etha-
In this moisture susceptibility study of asphalt mixtures, the nol were mixed at a ratio of 9:1 to dissolve the remaining part of the asphalt mix-
NHL was used to improve the moisture damage in the asphalt mix- ture samples. The non-polar and polar parts of the asphalt were obtained through
tures. The polar groups in the NHL modified and control asphalt these steps.
binder were extracted from the damaged asphalt mixtures and
analyzed. The aging effect of asphalt was evaluated on the water 3.3. Polar group analysis

damage in asphalt mixtures. The fundamental mechanism of mois- Infrared spectrometry (IR) is a useful tool in analyzing the chemical composi-
ture damage in the asphalt mixtures was explored with the MD tion and molecular bonds of materials. The Fourier Transform Infrared Spectroscopy
simulation. (FTIR) is a piece of technical equipment used for detecting the infrared absorption of
538 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547

13.5 cm

a b

Fig. 1. Appearance and FE-SEM image of nano hydrated lime materials: (a) white powder, (b) the FE-SEM image at the magnitude of 1000.

tested materials with a wide range of wavelengths. The FTIR attenuated total reflec- lime modified asphalt mixtures increases slightly. The moisture
tion (ATR) was used to test different types of samples [34]. The organic asphalt is
susceptibility of nano hydrated lime modified asphalt mixtures
composed of 90% carbon and hydrogen, and small amounts of other elements
[35]. It is known that six movements in the CH2 groups are common in organic
improves relative to the control asphalt binder from the TSR data.
materials, such as symmetrical stretching, rocking, anti-symmetrical stretching, The wet tensile strength of the control asphalt mixture decreases
twisting, scissoring, and wagging [36]. Of these movements, the symmetrical and faster than that of hydrated lime modified asphalt mixtures.
anti-symmetrical stretching movements are most common in asphalt. The Jasco Meanwhile, the dry tensile strength at 3 days for the hydrated lime
IRT 3000 FTIR spectrometer was employed to test the samples. The non-polar and
modified asphalt mixtures increases compared to the dry tensile
polar parts of asphalt were tested by FTIR ATR, and the spectra of samples were
analyzed. strengths at 2 days. This indicates that the addition of the nano
hydrated lime delays the moisture damage of modified asphalt
mixtures. The calcium ions from the NHL can prevent the displace-
4. Moisture susceptibility test of NHL modified asphalt mixtures ment of water at the asphalt binder–aggregate interfaces. Seen
from the reference [15], if the hydrated lime was blended with
The NHL was slowly added to the PG 58-28 asphalt and the the asphalt prior to the asphalt mixture, it can effectively reduce
modified asphalt was mixed in the high shear mixer at 135 °C the moisture damage of modified asphalt mixtures. In addition,
and 5000 rpm. Due to the high amount of NHL in the asphalt, the the trends of mixing and compaction temperatures of the control
shear speed was increased slightly when the sample was blended. and modified asphalt mixtures were followed by the temperatures
The modified asphalt binders were prepared for the asphalt mix- of the control asphalt mixture. If the mixing and compaction tem-
ture after about two hours of mixing. The aggregates were also peratures and the oxidation time of modified asphalt increases, the
pre-heated for approximately two hours to dry. The gradation strengths of modified asphalt mixtures increase more, as expected.
5E3 of aggregates was also used to compact the asphalt mixtures This also explains the low values for strengths of the dry and wet
by a SGC. The mixing and compaction temperatures were 150 NHL modified asphalt mixtures compared to the control mixture.
and 135 °C, respectively. Mixing was followed by the Superpave
compaction, and the loose asphalt mixtures were heated in the
oven (135 °C) for at least two hours. The samples were cured at 5. FTIR sample preparation and characterization results
room temperature for at least 72 h after the asphalt mixtures were
compacted. The asphalt mixture samples were tested by UTM and the
When the asphalt mixtures were ready, the indirect tensile test tensile strength of the asphalt mixtures was measured. The tested
was conducted and the tensile strength ratios (TSRs) of the control samples were dried at room temperature. Then, a small amount
and modified asphalt mixtures were evaluated. The test was based (around 20 g) of the tested samples was soaked in cyclohexane
on AASHTO’s procedure T283. The air voids were maintained solution (around 100 ml). The non-polar compounds in the asphalt
between 6% and 8%. The height of the sample was about melted in the cyclohexane. The first extraction of the asphalt
63.5 mm and the diameter was about 100 mm. The thaw cycles
were applied to the asphalt mixtures by immersing the sample
in the 60 °C water for different periods of time. All of the samples
were measured, using the Universal Testing Machine (UTM), for 1.2E+06 0.4
Dry and Moisture Tensile Strength (Pa)

the indirect tensile strength and its ratio under the conditions of Dry Tensile Strength (Pa) Moist Tensile Strenth (Pa) TSR

the room temperature and constant loading rate. These parameters 1.0E+06
Tensile Strength Ratio (%)

were calculated by Eq. (1), and the results are shown in Fig. 2.
8.0E+05

2P S2
St ¼ ; TSR ¼ ð1Þ 6.0E+05 0.2
ptD S1
4.0E+05
where St is the tensile strength (Pa), P is the maximum load (N), t is
the sample thickness (mm), D is the sample diameter (mm), TSR is 2.0E+05
the tensile strength ratio, S1 is the average dry tensile strength, and
0.0E+00 0
S2 is the average conditioned tensile strength. Control-2 days 10% Nano 20% Nano Control-3 days 10% Nano 20% Nano
Fig. 2 shows the tensile strengths of the control and NHL Hydrated Hydrated Hydrated Hydrated
Lime - 2 days Lime - 2 days Lime - 3 days Lime - 3 days
modified asphalt mixtures. The results show that the addition of
NHL into the asphalt mixtures decreases the dry tensile strength,
as well as the wet tensile strength. The results also coincide with Fig. 2. Tensile strength ratio (TSR) of the control and NHL modified asphalt
the trend of the reference [15]. However, the TSR of nano hydrated mixtures.
 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547 539

mixtures was done after the solution was mixed for about half an (C@O) of esters is found in a region of 1735–1750 cm1 from
hour and settled for a half hour. The second extraction of the asphalt Fig. 3a in the polar part of the asphalt. The peak of the car-
mixtures was performed using the solvent (around 40 ml) mixed bon–oxygen single bond (C–O) in the esters can be detected
with toluene–ethanol at a ratio of 9:1. The polar components in in an area of 1000–1300 cm1 from Fig. 3b. Due to the small
the asphalt were absorbed by the solvent, not by the cyclohexane. amount of esters in the oxidized asphalt, there is a limited
The solution was also mixed for about half an hour and settled for contribution to the aging of the asphalt [2].
a half hour. The weights of the samples and solutions were recorded, 6) Fig. 3a reveals that the peak of carbonyl groups (C@O) in the
as well as the non-polar and polar parts of the asphalt. The weight of amide is observed at a range of 1640–1690 cm1.
the polar components of the asphalt comprised less than 2% of the Fig. 3b and c also show that the N–H stretching and bending
weight of the tested asphalt mixtures. bonds are found in a region of 3100–3500 cm1 (including a
shoulder band) and 1550–1640 cm1, respectively, in the
5.1. Functional components in the asphalt asphalt at the liquid state. Due to the small amount of amide
[2] in the asphalt, the amide may not be easily detected.
The polar components of the control asphalt were tested at the
liquid state by the FTIR ATR. The spectrum data of the polar parts 5.2. FTIR results of polar groups in the control and NHL modified
of the control asphalt was analyzed to show the number of carbonyl asphalt
groups in the asphalt. The evidence of six functional groups with a
carbonyl group was found from the spectral images and references. After identifying the functional groups in the polar parts of the
The polar parts of the asphalt in the solution were examined by FTIR control asphalt, the liquid asphalt samples were dropped into the
ATR, and the FTIR results are shown in Fig. 3. In addition, the carbon– groove of FTIR and tested by FTIR ATR. The FTIR spectra were
oxygen (C–O) single bond can be observed in the ethanol spectrum recorded with the accumulation of 256 scans, with a resolution
[37–39]. One of the solutions used in the FTIR ATR test was mixed of 4 cm1. The peak range of functional polar carbonyl groups in
with toluene and ethanol, and the bonding features of toluene and the asphalt is demonstrated in Table 1. The polar compounds were
ethanol are shown in the spectra of the solution (Fig. 3). The features tested by the FTIR ATR at liquid state. The results of the FTIR spec-
of six functional groups with carbonyl are as follows. tra are shown in Fig. 4. The carbonyl groups in the polar part of the
asphalt were analyzed, and these groups may relate to the perfor-
1) There are different types of ketones, such as acyclic, cyclic, a, mance of the asphalt mixtures [15].
b-unsaturated, and aryl ketone. The ketones can be detected Fig. 4 shows the FTIR spectra of polar compounds in the
at a range of 1665–1850 cm1 [40,41], shown in Fig. 3a. asphalt at the liquid state. The solvents used were also tested
Ketones can reinforce the intensity of the carbonyl group with the FTIR ATR. The FTIR spectra of the control and NHL mod-
in asphalt at around 1700 cm1 [2]. They also have a close ified asphalt were compared, and the polar compositions in the
relationship to the performance of asphalt binders and mix- asphalt were analyzed. Based on the literature review [15–17],
tures [15–17,42]. the hydrated lime can enhance the adhesion between the aggre-
2) Three bonds, carbon–oxygen double stretching bond (C@O), gate and asphalt. The hydrated lime particles melted in the
carbon–oxygen single stretching bond (C–O), and oxygen– asphalt and the carboxylic acids/carbonyl groups in the asphalt
hydrogen bond (O–H), are observed in carboxylic acid improved the moisture susceptibility of the asphalt mixtures
[40,41]. Fig. 3a shows that the C@O peak of carboxylic acids including the mixture with siliceous aggregates. Compared to
is detected at a range of 1700–1725 cm1 [7] in the polar the spectra of the solutions, the peaks of the functional polar
groups of asphalt. Fig. 3b and c demonstrates the peaks of groups in asphalt were clearly displayed (section: functional
the carbon–oxygen single stretching bond (C–O) and hydro- group components). The ratios of carbonyl groups in the asphalt
xyl group (O–H) are observed at a range of 1210–1320 and were calculated using Eq. (2) through Eq. (8). The results of the
2500–3300 cm1, respectively. The carboxylic acids in the ratio calculations of the control and NHL modified asphalt at the
asphalt were also studied by Campbell [2], and the results liquid state are shown in Fig. 5.
show that carboxylic acids are also aging products of
asphalt. Area of the carboxylic acid bands between 1720 and 1700cm1
Iacids ¼ P
3) Fig. 3a shows that two main carbonyl groups are detected at Area of the spectral bands between 2000 and 600cm1
the bands [40,41] around 1800–1830 and 1740–1775 cm1, ð2Þ
respectively. The anhydrides are also produced during the
oxidation of asphalt and the two carbonyl groups observed Area of the aldehyde bands between 1735 and 1720cm1
Ialdehydes ¼ P
should be anhydrides. A limited anhydrides were observed Area of the spectral bands between 2000 and 600cm1
during oxidation, but large amounts of anhydrides were ð3Þ
detected in the long-term aged asphalt [43]. Stronger absor-
bance appeared at a range of 1740–1775 cm1, and this peak Area of the amide bands between 1660 and 1645cm1
Iamides ¼ P ð4Þ
area was used to represent the anhydrides in this study. Area of the spectral bands between 2000 and 600cm1
4) The Refs. [40,41] display that there are two obvious bonds,
carbonyl (C@O) and carbon–hydrogen bonds (C–H), in the Area of the anhydride bands between 1775 and 1750cm1
aldehydes. The peak of the carbonyl group in the aldehyde Ianhydrides ¼ P
Area of the spectral bands between 2000 and 600cm1
is detected at a range of 1720–1740 cm1, as shown in ð5Þ
Fig. 3a. It is easy to identify two peaks of carbon–hydrogen
bonds of organic materials at a range of 2820–2850 and Area of the ester bands between 1750 and 1735cm1
2720–2750 cm1 in Fig. 3c. Campbell studied asphalt oxida- Iesters ¼ P ð6Þ
Area of the spectral bands between 2000 and 600cm1
tion, and the results reveal that the aldehydes also increased
the carbonyl absorbance in the oxidized asphalt [2]. Area of the aryl ketone bands between 1700 and 1685cm1
5) The carbon–oxygen double stretching bond (C@O) and car- Iaryl ketones ¼ P
Area of the spectral bands between 2000 and 600cm1
bon–oxygen single stretching bond (C–O) are observed in
ð7Þ
the ester group [40,41]. The peak of the carbonyl group
540 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547

Control polar asphalt

0.07
Control polar asphalt
Solution

Carboxylic acid
0.06 Ester

Absorbance(a.u.)
Amide
Anhydride Aldehyde Ketone
Anhydride

0.05

0.04
1900 1850 1800 1750 1700 1650 1600
Wavenumber(cm-1)

(a) Functional groups containing the carbonyl group in the polar asphalt (liquid state)

Control polar asphalt

0.20
Control polar asphalt
Solution

0.16
Absorbance(a.u.)

0.12
Nitrogen-Hydrogen
(N-H) bending area

Carbon-Oxygen
(C-O) stretching area
0.08

0.04
1700 1600 1500 1400 1300 1200 1100 1000 900
Wavenumber(cm-1)

(b) Nitrogen-hydrogen and carbon-oxygen bonds in the polar asphalt (liquid state)

(c) Oxygen-hydrogen and Nitrogen-hydrogen bonds in the polar asphalt (liquid state)

Fig. 3. Chemical bonds in the asphalt binders.

 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547 541

Table 1 groups (anhydrides, ketones and carboxylic acids) in the oxidized

Functional groups containing a carbonyl in the asphalt binder. asphalt. Seen from Fig. 5a, the amounts of functional carbonyl
Functional groups Absorption frequency region, wavenumber (cm1) groups in the 20% NHL modified asphalt are more than the control
Carbonyl: C@O 1670–1820, stretching vibration, strong intensity and the 10% NHL modified asphalt. The quantity of carbonyl groups
[40,41,44] in the 10% NHL modified asphalt is slightly more than that of the
Functional groups: carboxylic acid control asphalt. The addition of NHL into the control asphalt
C@O 1700–1725, stretching vibration [40,41,44] increases the functional groups in the modified asphalt in the dry
O–H 2500–3300, stretching vibration, strong intensity condition.
[40,41,44] Fig. 5b and c reveals the ratios of the functional groups in the
C–O 1210–1320, stretching vibration, strong intensity
[40,41,44]
control and NHL modified asphalt after 2- and 3-day wet condi-
tions. The amount of carbonyl groups in the 20% NHL modified
Functional groups: aldehyde
C@O 1720–1740, stretching vibration, strong intensity
asphalt is more than those in the control and 10% NHL modified
[40,41,44] asphalt, and the quantity of carbonyl groups in the 10% NHL
@C–H 2820–2850 and 2720–2750, stretching vibration, modified asphalt is more than that of the control asphalt under
medium intensity [40,41,44] the wet condition. The anhydrides, carboxylic acids and ketones
Functional groups: amide are still the leading products during the aging process in the polar
C@O 1640–1690, stretching vibration, strong intensity asphalt at a liquid state. Considering the TSR results of the control
[40,41,44]
and NHL modified asphalt mixtures, the more polar groups that are
N–H 3100–3500, stretching vibration, unsubstituted [40,41]
N–H 1550–1640, bending vibration [40,41] in the asphalt, the better the TSR results are. This indicates that
asphalt aging under the wet condition helps the asphalt mixture
Functional groups: ketone
Acyclic 1705–1725, stretching vibration, strong intensity to resist moisture damage, especially the ketones, carboxylic acids,
[40,41] and anhydride in this case. The calcium ions in the NHL modified
Cyclic 3-Membered-1850, stretching vibration, strong asphalt may be the reason for the improvement of the moisture
intensity [40,41] damage in the asphalt mixtures.
4-Membered-1780, stretching vibration, strong
intensity [40,41]
Therefore, the ketones and carboxylic acids increased more
5-Membered-1745, stretching vibration, strong after the oxidation of asphalt, and the anhydrides and amides
intensity [40,41] increased, in this case. The aging groups in the asphalt enhance
6-Membered-1715, stretching vibration, strong the resistance to permanent deformation in the asphalt pavement
intensity [40,41]
at an earlier stage and also help to improve the moisture suscepti-
7-Membered-1705, stretching vibration, strong
intensity [40,41] bility of asphalt mixtures. Furthermore, the effect of NHL materials
a, b-unsaturated 1665–1685, stretching vibration, strong intensity on the resistance to moisture damage in asphalt mixtures is
[40,41] positive.
Aryl ketone 1680–1700, stretching vibration, strong intensity
[40,41]
Functional groups: anhydride 6. Molecular dynamics (MD) analysis
C@O 1800–1830 and 1740–1775, stretching vibration
[40,41,44]
The bonding strength between the aggregates and the asphalt
Functional groups: ester determines the performance of the rutting resistance and moisture
C@O 1735–1750, stretching vibration, strong intensity
susceptibility in asphalt mixtures [9,45]. The mechanism of water
[40,41,44]
C–O 1000–1300, stretching vibration, two bands or more replacement in the moisture damage in asphalt mixtures is
[40,41,44] demonstrated in Fig. 6, from the initiation to the propagation stage.
The bonding strength is related to the adhesion energy between

Area of the a; b unsaturated ketone bands between 1680 and 1665 cm1
Ia;b unsaturated ketones ¼ P ð8Þ
Area of the spectral bands between 2000 and 600 cm1

During the FTIR sample preparations, it is possible that the con- the asphalt and aggregates, which varies with the interfacial
centrations of polar groups in the asphalt were affected by the interactions among asphalt, water, and aggregates. Therefore, the
solutions. However, the trends of these functional groups were adhesion energy calculation is important to explain the mecha-
not changed. It is likely that the composition produced by the nism of moisture distresses in the asphalt mixtures. In this study,
asphalt oxidation may be changed by different conditions. Fig. 5a the interaction between the asphalt and aggregates, as well as
shows the ratios of functional polar groups in the control and the interaction between the water and aggregates, was first simu-
NHL modified asphalt at the dry condition. The amount of anhy- lated using the molecular dynamics method. Through MD simula-
dride is the highest in the polar groups in the asphalt. If the sum tions of the potential energies of the interface and individual
of aryl ketones and a,b-unsaturated ketones are considered as models, the adhesion energies were computed to explain the cause
ketones, then, the ketones are the main products in the functional of moisture damage in asphalt mixtures. The aging effects of
polar groups of asphalt. After the asphalt mixture tests, it is deter- asphalt binders were demonstrated for the mitigation of water
mined that the carboxylic acids in the polar groups of asphalt are damage in asphalt mixtures.
also a major part of the asphalt oxidation products. The amount When the asphalt and aggregates are mixed, the aggregates are
of esters, aldehydes and amides is less than the three primary heated in the oven for at least two hours. It is assumed that water
542 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547

Control polar asphalt

0.14
Control polar asphalt - dry condition
0.13 Control polar asphalt - 2-day wet condition
Control polar aspahlt - 3-day wet condition
0.12 Solution

0.11

Absorbance(a.u.)
0.1

0.09

0.08

0.07

0.06

0.05

0.04
2000 1900 1800 1700 1600 1500
-1
Wavenumber(cm )

(a) FTIR spectra of the polar parts in the control asphalt at a range of 1500-2000cm-1
10% nano hydrated lime modified polar asphalt
0.17
10% nano hydrated lime modified polar asphalt - dry condition
0.16 10% nano hydrated lime modified polar asphalt - 2-day wet condition
0.15 10% nano hydrated lime modified polar asphalt - 3-day wet condition
solution
0.14
0.13
Absorbance(a.u.)

0.12
0.11
0.10
0.09
0.08
0.07
0.06
0.05
0.04
2000 1900 1800 1700 1600 1500
-1
Wavenumber(cm )

(b) FTIR Spectra of the polar parts in 10% NHL modified asphalt at a range of 1500-2000cm-1

20% nano hydrated lime modified polar asphalt

0.16
20% nano hydrated lime modified polar asphalt - dry condition
0.15 20% nano hydrated lime modified polar asphalt - 2-day wet condition
20% nano hydrated lime modified polar asphalt - 3-day wet condition
0.14
solution
0.13
0.12
Absorbance(a.u.)

0.11
0.10
0.09
0.08
0.07
0.06
0.05
0.04
2000 1900 1800 1700 1600 1500
-1
Wavenumber(cm )

(c) FTIR spectra of the polar parts in 20% NHL modified asphalt at the range of 1500-2000cm-1

Fig. 4. FTIR results of polar parts in the asphalt binders.

on the surface or in the porous holes of the aggregates is vaporized mental conditions decrease the strength of adhesion between the
and escapes from the aggregates. This confirms that the water aggregates and asphalt/asphalt mastic and the cohesive bonding
inducing the damage in the asphalt mixtures is from the external strength within the asphalt binder [46,47]. It is also assumed that
environment. The repeated traffic loads and changes in environ- the asphalt mastic consists of the asphalt binder and fine
 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547 543

0.014 In these interface models, the crystal quartz (SiO2) structure is

Control under dry condition 10% NHL under dry condition
20% NHL under dry condition represented as the aggregate in the asphalt mixtures. Based on
0.012
the Corbett method [35], the asphalt material is composed of
0.01 asphaltenes, polar aromatics, naphthene aromatics, and saturates
Ratio of bonds

through different absorption and desorption. Based on the refer-

0.008
ences [50,51], the asphalt model consists of asphaltenes, saturates,
0.006 and aromatics at the ratio of 5:41:27, respectively. The structure of
asphaltenes in the asphalt model was proposed by Artok et al. [52]
0.004
and Zhang et al. [53]. These saturates were represented by
0.002 docosane molecules and the aromatics were replaced by the
1,7-dimethylnaphthalene in the asphalt model from the references
0
Acid Aldehyde Amide Anhydride Ester Aryl Ketone αβ-ketone
[53–56]. In addition, based on the structure of asphalt components
1700-1720 1720-1735 1645-1660 1750-1775 1735-1750 1685-1700 1665-1680 from the references [53], the aging effects on these components
were investigated and one of the aging groups (the carboxylic acid
(a) Ratio calculations of functional polar groups in the control and NHL modified asphalt at
the dry condition (liquid state) group) was attached in the structure of the components (Fig. 7).
The possible structures of asphaltenes (Fig. 7a) and aromatics
0.014
Control under 2-day wet conditon 10% NHL under 2-day wet condition (Fig. 7b) with the carboxylic acid group are proposed and displayed
20% NHL under 2-day wet condition
0.012 in Fig. 7. However, the structure of docosane (Fig. 7c) is relatively
stable, and it is likely that the docosane is not compatible with
0.01
other oxidizing agents.
Ratio of bonds

0.008 Three interface models among the asphalt, water, and aggre-
gates were created by the Materials Processes and Simulations
0.006
(MAPS) software with the Amber Cornell Extension Force Field
0.004 [57,58]. These interface models include the interaction systems
of (1) asphalt–aggregate, (2) aggregate–water, and (3) aggregate–
0.002
asphalt with carboxylic acid (aging group) (Fig. 8). The density of
0 asphalt in the interface model is around 1.04 g/cm3, which matches
Acid Aldehyde Amide Anhydride Ester Aryl Ketone αβ-ketone
the physical properties of asphalt. Parts of bond parameters of the
1700-1720 1720-1735 1645-1660 1750-1775 1735-1750 1685-1700 1665-1680
Amber Cornell Extension Force Field are from the Amber Cornell
(b) Ratio calculations of functional polar groups in the control and NHL modified asphalt Force Field and General Amber Force Field (GAFF) [57,58]. The
after a 2-day wet condition (liquid state) system energy formula in the Amber Cornell Extension Force Field
0.014 is shown in Eq. (9).
Control under 3-day wet condition 10% NHL under 3-day wet condition
20% NHL under 3-day wet condition X X
0.012 Etotal ¼ K r ðr  r eq Þ2 þ K h ðh  heq Þ2
bonds angles
0.01 " #
X Vn X Aij Bij qi qj
Ratio of bonds

0.008 þ ½1 þ cos ðnu  cÞ þ  6þ ð9Þ

dihedrals
2 12
i<j Rij Rij 2 Rij
0.006
where the first term is the energy between the covalently bonded
0.004
atoms; the second term is the energy caused by the geometry of
0.002 electron orbitals; the third term is the energy from twisting a bond;
the fourth term is a non-bonded energy from van der Waals and
0
Acid Aldehyde Amide Anhydride Ester Aryl Ketone αβ-ketone electrostatic energies (atom pairs). r eq and heq are the equilibrium
1700-1720 1720-1735 1645-1660 1750-1775 1735-1750 1685-1700 1665-1680 structural parameters; K r and K h are the force constants; n is the
(c) Ratio calculations of functional polar groups in the control and NHL modified asphalt multiplicity and c is the phase angle for the torsional angle param-
after a 3-day wet condition (liquid state) eters; A, B and q are the non-bonded potentials between all atom
Fig. 5. Ratio calculations of functional polar groups in the control and NHL modified
pairs; and finally, Rij is the distance of atoms and 2 is the well depth
asphalt. for van der Waals energy.
In the interaction models, the Electrostatic Potential (ESP)
charges were allotted to the asphalt components by using the
Asphalt quantum mechanics module NWChem in the MAPS software
Asphalt Asphalt
[59]. The Conjugate Gradient method and the Particle Mesh Ewald
Aggregates Aggregates (PME) method were used to optimize the energy of the system. The
Aggregates
Conjugate Gradient method was used to optimize the energy in the
Water large and complex systems. The optimization calculation formula
is shown in Eq. (10).
Asphalt-aggreates Water Initiation Propagation
1 T
Fig. 6. Stages of water displacement in the moisture damage of asphalt mixtures. f ðxÞ ¼ x Ax  xT b; x 2 Rn ; ðiterationsÞ ð10Þ
2
aggregates (aggregate size is normally smaller than 2.36 mm), and where A is symmetric, positive and real, b is a known coefficient.
the asphalt mastic is in the transition state to form the asphalt PME is an Ewald summation for the long-range potential com-
mixtures [48,49]. When the adhesive and cohesive bonds are putation in the periodic systems. In the PME method, the long-
reduced in the asphalt mixtures, water starts to intrude into the range part is divided into two sections: short-range and long-
aggregates, and moisture damage in asphalt mixtures develops range contributions. The calculations of two-part energies are
from the initiation stage to the propagation stage (Fig. 6). expressed in Eq. (11).
544 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547

Asphaltene with the

carboxylic acid group

Sum Formula: C64H50O2S2

(a) asphaltene structure with the carboxylic acid group (aging group)

Sum Formula: C12H10O2

c
b
Sum Formula: C22H46

(b) 1, 7-dimethylnaphthalene with the carboxylic acid group (aging group)

(c) Docosane structure, represents a saturate.

Fig. 7. Three components in the asphalt model with the carboxylic acid group (aging group). Note: C–C symbol connection shows the double bond of carbon atoms in a
benzene ring; No C–C symbol connection represents a single bond of carbon atoms.

uðrÞ def
¼ usr ðrÞ þ ulr ðrÞ ð11Þ energy of the substrate system, and aggregate system (kcal/mol); A
is the area of each interface system (Angstrom2).
where usr ðrÞ is the short-range term potential energy which quickly To calculate the adhesive energy for the asphalt–aggregate, an
converges in the real space, and ulr ðrÞ is the long-range term poten- interface model with a specific vacuum layer and periodic bound-
tial which quickly converges in the Fourier space. ary condition was applied as shown in Fig. 8a. The aggregate model
Fig. 8 shows the interface MD models of asphalt–aggregate, includes a surface layer with the hydroxyl group (–OH) for the fol-
water–aggregate, aggregate–asphalt with the carboxylic acid lowing study. The separate aggregate models with the surface
group systems. The adhesion energy of each system directly relates function group (–OH) and asphalt model were developed as shown
to the bonding strength of the system. The adhesion energy of each in Fig. 8b and c. The NVT simulation was conducted in three inter-
interface system is computed using Eq. (12). face systems and in each separate system. The potential energies of
Einterface  ðEtop þ Esubstrate Þ these systems were computed and outputted. The energy conver-
Eadhesion ¼ ð12Þ sion factor was used to present the adhesion energy (conversion
A
factor: 1 kcal
mol
¼ 6:95  1021 J). Avogadro’s number was used to con-
where Eadhesion is the adhesion energy of the interface system
vert the energy unit. After computing and unit conversion, the
(kcal/mol); Einterface is the potential energy of the interface
adhesion energy of the asphalt–aggregate system is
system (kcal/mol); Etop is the potential energy of the top layer
0.1446295 J/m2. Similarly, the adhesion energy calculation of
system, asphalt, or water system (kcal/mol); Esubstrate is the potential
 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547 545

a b c

1) The molecular structures of the asphalt-aggregate system: (a) aggregate-asphalt system; (b)
only the asphalt system; and (c) only the aggregate system with functional group OH on the
surface layer.

d e f

2) The molecular structures of the aggregate-water system: (d) aggregate-water system; (e) only
the water system; and (f) only the aggregate system with functional group OH on the surface
layer.

Carboxylic Carboxylic
acid g acid h i

3) The molecular structures of the asphalt-aggregate system with aging groups: (a) aggregate-
asphalt system with the carboxylic acid group; (b) only the asphalt system with the
carboxylic acid group; and (c) only the aggregate system with functional group OH on the
surface layer.

Fig. 8. MD models for (1) asphalt–aggregate interactions; (2) water–aggregate interactions and (3) aggregate–asphalt with the carboxylic acid group systems.

the water–aggregate interface was conducted with the interface gate–water system was calculated as 1.86955 J/m2. In order to
model and two separate systems of the water layer and aggregate demonstrate the effects of aging on the asphalt, on the adhesion
substrate, shown in Fig. 8d–f. The adhesion energy of the aggre- energy interface, the interface model of the aggregate–asphalt with
546 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547

the carboxylic acid group and two separate systems were reason for moisture damage in asphalt mixtures. The water
developed as shown in Fig. 8g–i. The adhesion energy of the aggre- tends to adhere to aggregates rather than to asphalt under
gate–asphalt with the carboxylic acid group system was evaluated the same condition. In addition, from investigating the dif-
as 0.151371 J/m2. ferent adhesion energies between the aggregate and asphalt
The different adhesion energies of asphalt–aggregate and aggre- with or without the carboxylic acid group (aging group), the
gate–water systems indicate that the bonding strength between moderately aged asphalt improves the energy of adhesion
the asphalt and aggregate is lower than the attraction force and, thus, helps resist moisture damage in asphalt mixtures.
between aggregate and water. The energy difference is also the
reason why water bonds to the aggregates easier than the asphalt With chemical extraction, the polar and non-polar components
in asphalt mixtures. Therefore, coating the aggregates with asphalt in the asphalt were separated and analyzed to link the moisture
is the key process in making asphalt mixture samples or paving susceptibility of asphalt mixtures with FTIR characterization.
asphalt pavement. It is also very important to dry the aggregates Ketones, carboxylic acids and anhydrides are the main compounds
prior to mixing. These procedures make sure that water completely present during asphalt oxidation in this study, and these also relate
vaporizes from the aggregate pores and surfaces. These are also to the performance of the pavement. MD simulations discovered
preventive measures/practices to reduce the moisture damage in the adhesion energy difference between asphalt–aggregate and
asphalt mixtures. Furthermore, the energy of adhesion of the aggregate–water systems. The analysis results explain the
aggregate–asphalt (with the carboxylic acid group) system is chemo-physical properties of asphalt related to the moisture
greater than that of the asphalt–aggregate system. The varying damage in asphalt mixtures. In addition, the future research of
energies of two systems imply that the presence of the aging group our group will focus on simulating asphalt aging.
in asphalt causes a stronger adhesion force between the asphalt
and aggregate. This phenomenon also reasonably interprets that
studying the mildly aged asphalt is helpful in improving the water Acknowledgments
susceptibility of asphalt mixtures.
The authors would like to thank Dr. Yoke Khin Yap and Mingx-
7. Discussions and conclusions iao Ye for the help on FTIR testing. The authors also appreciate the
help of Dr. Andreas Bick on generating the molecular dynamics
The control and NHL modified asphalt were tested by FTIR ATR model. The experimental work was completed at the Transporta-
to analyze the polar groups with carbonyl in the asphalt at a liquid tion Materials Research Center of Michigan Technological Univer-
state. Parts of the control and the NHL modified asphalt mixture sity. The authors appreciate the financial support of the U.S.
samples tested by TSR were also used to be extracted by solutions. National Science Foundation (NSF – United States) under the Grant
These solutions contained the polar functional groups that were number: 1300286. Any opinion, finding, and conclusion or recom-
detected by the FTIR ATR to analyze the carbonyl groups in the mendation expressed in this material are those of the authors and
asphalt. The mechanical driving forces for water damage between do not necessarily reflect the view of any organization.
the asphalt–aggregate and aggregate–water systems in asphalt
mixtures were explored and discussed using molecular dynamics
simulations. Based on the data analysis of the extracted control References
and the NHL modified asphalt, the following conclusions can be
[1] J.C. Petersen, P.M. Harnsberger, R.E. Robertson, Factors affecting the kinetics
drawn. and mechanisms of asphalt oxidation and the relative effects of oxidation
products on age hardening, Am. Chem. Soc. Div. Fuel Chem. 41 (4) (1996)
1) During the oxidation of asphalt, the resins and maltenes in 1232–1244.
[2] P.G. Campbell, J.R. Wright, Infrared spectra of asphalts: some aspects of the
the asphalt helped to generate polar group components,
changes caused by photooxidation, J. Res. Natl. Bur. Std. Sect. C 68 (2) (1964)
including carbonyl groups, sulfoxide groups, and nitrogen 115–123.
oxides. Carboxylic acids and ketones were the major [3] W.D. Fernandez-Gomez, H.R. Quintana, F.R. Lizcano, A review of asphalt and
asphalt mixture aging, Ingeniería e Investigación 33 (1) (2013) 5–12.
products of the oxidized asphalt, which also produces high
[4] S. Caro, A. Diaz, D. Rojas, H. Nuñez, A micromechanical model to evaluate the
amounts of anhydrides. The limited amount of amides, impact of air void content and connectivity in the oxidation of asphalt
esters, and aldehydes in the oxidized and extracted asphalt mixtures, Constr. Build. Mater. 61 (2014) 181–190.
was detected. Carboxylic acids, ketones and anhydrides are [5] P.E. Yuhong Wang, K. Zhao, C. Glover, L. Chen, Y. Wen, D. Chong, et al., Effects of
aging on the properties of asphalt at the nanoscale, Constr. Build. Mater. 80
the pivotal components in the asphalt that link to the rutting (2015) 244–254.
and moisture resistance in asphalt mixtures, as well as the [6] L. Xiang, J. Cheng, S. Kang, Thermal oxidative aging mechanism of crumb
bonding strength between the asphalt and aggregates. The rubber/SBS composite modified asphalt, Constr. Build. Mater. 75 (2015) 169–
175.
aging of asphalt definitely improves the resistance to mois- [7] H. Yao, Z. You, L. Li, S.W. Goh, C.H. Lee, Y.K. Yap, et al., Rheological properties
ture damage in wet environment at an early stage according and chemical analysis of nanoclay and carbon microfiber modified asphalt
to these research findings. However, the long-term aging of with Fourier transform infrared spectroscopy, Constr. Build. Mater. 38 (2013)
327–337.
asphalt causes the degradation of the bonding strength [8] Z. You, J. Mills-Beale, J.M. Foley, S. Roy, G.M. Odegard, Q. Dai, et al., Nanoclay-
between the asphalt and aggregates, as well as the fatigue modified asphalt materials: preparation and characterization, Constr. Build.
life and resistance to moisture. Mater. 25 (2) (2011) 1072–1078.
[9] H. Yao, Z. You, L. Li, X. Shi, S.W. Goh, J. Mills-Beale, et al., Performance of
2) The interface MD models of asphalt–aggregate and
asphalt binder blended with non-modified and polymer-modified nanoclay,
aggregate–water systems were generated to simulate the Constr. Build. Mater. 35 (2012) 159–170.
mechanism that produces adhesive energy, and also to [10] S.W. Goh, M. Akin, Z. You, X. Shi, Effect of deicing solutions on the tensile
strength of micro- or nano-modified asphalt mixture, Constr. Build. Mater. 25
mimic the interactions between aggregates and asphalt/
(1) (2011) 195–200.
water. The potential energies of each MD interface system [11] S. Zhao, B. Huang, X. Shu, M. Woods, Comparative evaluation of warm mix
and separated systems (asphalt, water, and aggregate) were asphalt containing high percentages of reclaimed asphalt pavement, Constr.
computed to determine the bonding energy of each system. Build. Mater. 44 (2013) 92–100.
[12] S. Zhao, B. Huang, X. Shu, X. Jia, M. Woods, Laboratory performance evaluation
The difference of adhesive energies between the asphalt– of warm-mix asphalt containing high percentages of reclaimed asphalt
aggregate and aggregate–water systems is the fundamental pavement, Transp. Res. Rec. J. Transp. Res. Board 2294 (2012) 98–105.
 H. Yao et al. / Construction and Building Materials 101 (2015) 536–547 547

[13] X. Shu, B. Huang, E.D. Shrum, X. Jia, Laboratory evaluation of moisture [37] NIST Mass Spec Data Center SES, director, Infrared spectra, NIST Chemistry
susceptibility of foamed warm mix asphalt containing high percentages of WebBook, NIST Standard Reference Database Number 69, National Institute of
RAP, Constr. Build. Mater. 35 (2012) 125–130. Standards and Technology, Gaithersburg, MD 20899, United States, 2014.
[14] J. Cheng, J. Shen, F. Xiao, Moisture susceptibility of warm-mix asphalt mixtures [38] R.A. Tomasi, A Spectrum of Spectra, Sunbelt R&T, Incorporated, 1992, p. 136.
containing nanosized hydrated lime, J. Mater. Civ. Eng. 23 (11) (2011) 1552– [39] R.A. Tomasi, A Spectrum of Spectral Problems: Supplement, Sunbelt R&T,
1559. Incorporated, 1994, p. 50.
[15] S. Huang, R. Robertson, J. Branthaver, J. Claine Petersen, Impact of lime [40] R.M. Silverstein, G.C. Bassler, Spectrometric identification of organic
modification of asphalt and freeze–thaw cycling on the asphalt–aggregate compounds, J. Chem. Educ. 39 (11) (1962) 546.
interaction and moisture resistance to moisture damage, J. Mater. Civ. Eng. 17 [41] R.M. Silverstein, G.C. Bassler, T.C. Morrill, Spectrometric Identification of
(6) (2005) 711–718. Organic Compounds, Wiley, New York, 1981, p. 442.
[16] S.-C. Huang, F.P. Miknis, W. Schuster, S. Salmans, R. Boysen, Rheological and [42] D.N. Little, J.A. Epps, P.E. Sebaaly, The Benefits of Hydrated Lime in Hot Mix
chemical properties of hydrated lime and polyphosphoric acid modified Asphalt, National Lime Association, 2006, p. 80.
asphalts with long term aging, J. Mater. Civ. Eng. Am. Soc. Civ. Eng. 23 (5) [43] J.C. Petersen, National Research C, Transportation Research B, Committee on
(2011) 628–637. Characteristics of Bituminous M, A Review of the Fundamentals of Asphalt
[17] D. Lesueur, J. Petit, H.-J. Ritter, The mechanisms of hydrated lime modification Oxidation Chemical, Physicochemical, Physical Property, and Durability
of asphalt mixtures: a state-of-the-art review, Road Mater. Pavement Des. 14 Relationships, 2009, p. 78.
(1) (2012) 1–16. [44] H. Yao, Z. You, L. Li, C. Lee, D. Wingard, Y. Yap, et al., Rheological properties and
[18] B. Huang, X. Shu, Q. Dong, J. Shen, Laboratory evaluation of moisture chemical bonding of asphalt modified with nanosilica, J. Mater. Civ. Eng. 25
susceptibility of hot-mix asphalt containing cementitious fillers, J. Mater. (11) (2013) 1619–1630.
Civ. Eng. 22 (7) (2010) 667–673. [45] H. Yao, L. Li, X.-L. Yang, H.-C. Dan, S. Luo, Mechanics performance research and
[19] I. Ishai, J. Craus, Effect of the filler on aggregate-bitumen adhesion properties microstructure analysis of nanomaterials modified asphalt, J. Build. Mater. 14
in bituminous mixtures, Association of Asphalt Paving Technologists Proc., (5) (2011) 712–717.
from the Technical Session, San Antonio, Texas, 21–23 February, 1977, pp. [46] S. Caro, E. Masad, A. Bhasin, D. Little, Coupled micromechanical model of moisture-
228–258. induced damage in asphalt mixtures, J. Mater. Civ. Eng. 22 (4) (2010) 380–388.
[20] J. Blazek, G. Sebor, D. Maxa, M. Ajib, H. Paniagua, Effect of hydrated lime [47] H. Yao, Y. Liu, Z. You, L. Li, S.W. Goh, Discrete element simulation of bending
addition on properties of asphalt, Pet. Coal 42 (1) (2000) 41–45. beam rheometer tests for asphalt binder, Int. J. Pavement Res. Technol. 5 (3)
[21] S. Plimpton, Fast parallel algorithms for short-range molecular dynamics, J. (2012) 161–168.
Comput. Phys. 117 (1) (1995) 1–19. [48] Y. Liu, Z. You, Q. Dai, J. Mills-Beale, Review of advances in understanding
[22] A. Bandyopadhyay, Molecular Modeling of EPON 862-DETDA Polymer, impacts of mix composition characteristics on asphalt concrete (AC)
Michigan Technological University, Houghton, MI, United States, 2012, p. 100. mechanics, Int. J. Pavement Eng. 12 (4) (2011) 385–405.
[23] Y. Lu, L. Wang, Nanoscale modelling of mechanical properties of asphalt– [49] S. Adhikari, Z. You, Investigating the sensitivity of aggregate size within sand
aggregate interface under tensile loading, Int. J. Pavement Eng. 11 (5) (2010) mastic by modeling the microstructure of an asphalt mixture, J. Mater. Civ.
393–401. Eng. 23 (5) (2011) 580–586.
[24] Chemistry IUoPaA, Compendium of Chemical Terminology, Version 2.3, [50] L. Zhang, M.L. Greenfield, Relaxation time, diffusion, and viscosity analysis of
International Union of Pure and Applied Chemistry, 2011, p. 1622. model asphalt systems using molecular simulation, J. Chem. Phys. 127 (19)
[25] R.T.B.R.N. Morrison, Organic chemistry, Englewood Cliffs, NJ, Prentice-Hall, (2007) 1–13.
1993, p. 1279. [51] L. Zhang, M.L. Greenfield, Molecular orientation in model asphalts using
[26] J. McMurry, Organic Chemistry, Cengage Learning, 2011, p. 1376. molecular simulation, Energy Fuels 21 (2) (2007) 1102–1111.
[27] H. Held, A. Rengstl, D. Mayer, Acetic anhydride and mixed fatty acid [52] L. Artok, Y. Su, Y. Hirose, M. Hosokawa, S. Murata, M. Nomura, Structure and
anhydrides, Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-VCH reactivity of petroleum-derived asphaltene, Energy Fuels 13 (2) (1999) 287–296.
Verlag GmbH & Co. KGaA, 2000, p. 150. [53] L. Zhang, M.L. Greenfield, Analyzing properties of model asphalts using
[28] F.A. Carey, R.J. Sundberg, Advanced Organic Chemistry: Part B: Reaction and molecular simulation, Energy Fuels 21 (3) (2007) 1712–1716.
Synthesis, Springer, 2007, p. 1270. [54] L. Zhang, M.L. Greenfield, Effects of polymer modification on properties and
[29] M. Jerry, Advanced Organic Chemistry: Reactions, Mechanisms and Structure, microstructure of model asphalt systems, Energy Fuels 22 (5) (2008) 3363–
1985, p. 1346. 3375.
[30] C.R. Kemnitz, M.J. Loewen, ‘‘Amide resonance” correlates with a breadth of C– [55] I. Kowalewski, M. Vandenbroucke, A.Y. Huc, M.J. Taylor, J.L. Faulon, Preliminary
N rotation barriers, J. Am. Chem. Soc. 129 (9) (2007) 2521–2528. results on molecular modeling of asphaltenes using structure elucidation
[31] W. Riemenschneider, H.M. Bolt, Esters, organic, Ullmann’s Encyclopedia of programs in conjunction with molecular simulation programs, Energy Fuels 10
Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, 2000, p. 150. (1) (1996) 97–107.
[32] R.J. Williams, A. Gabriel, R.C. Andrews, The relation between the hydrolysis [56] H. Groenzin, O.C. Mullins, Molecular size and structure of asphaltenes from
equilibrium constant of esters and the strengths of the corresponding acids, J. various sources, Energy Fuels 14 (3) (2000) 677–684.
Am. Chem. Soc. 50 (5) (1928) 1267–1271. [57] J. Wang, R.M. Wolf, J.W. Caldwell, P.A. Kollman, D.A. Case, Development and
[33] B. Neises, W. Steglich, Esterification of carboxylic acids with testing of a general amber force field, J. Comput. Chem. 25 (9) (2004) 1157–
dicyclohexylcarbodiimide/4-dimethylaminopyridine: tert-butyl ethyl 1174.
fumarate, Organic Syntheses, John Wiley & Sons, Inc., 2003, p. 183. [58] W.D. Cornell, P. Cieplak, C.I. Bayly, I.R. Gould, K.M. Merz, D.M. Ferguson, et al.,
[34] P. Griffiths, J.A. De Haseth, Fourier Transform Infrared Spectrometry, John A second generation force field for the simulation of proteins, nucleic acids,
Wiley & Sons, 2007, p. 560. and organic molecules, J. Am. Chem. Soc. 117 (19) (1995) 5179–5197.
[35] Asphalt Institute, Superpave performance graded asphalt binder specification [59] M. Valiev, E.J. Bylaska, N. Govind, K. Kowalski, T.P. Straatsma, H.J.J. Van Dam,
and testing, Superpave Series No 1 (SP-1), Lexington, KY, USA, 2003, p. 15. et al., NWChem: a comprehensive and scalable open-source solution for large
[36] P. Atkins, J. de Paula, Elements of Physical Chemistry, OUP Oxford, 2013, scale molecular simulations, Comput. Phys. Commun. 181 (9) (2010) 1477–
p. 648. 1489.