Construction and Building Materials 38 (2013) 759–765

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials

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

Preparation and properties of conventional asphalt modified by physical mixtures

of linear SBS and montmorillonite clay
Martin Jasso a,c,⇑, Dusan Bakos a, Daryl MacLeod b, Ludo Zanzotto c
a
Department of Plastics and Rubber, Institute of Polymer Materials, Faculty of Chemical and Food Technology, Slovak University of Technology, Radlinského 9,
812 37 Bratislava, Slovak Republic
b
Husky Energy, 707 8th Ave. SW, Calgary, Alberta, Canada T2P 3G7
c
Department of Civil Engineering, Schulich School of Engineering, University of Calgary, ENF 262, 2500 University Drive NW, Calgary, Alberta, Canada T2N 1N4

h i g h l i g h t s

" Asphalt blends studied via three factors and five levels of statistical experiment.
" The linear viscoelastic properties were studied at high temperatures.
" Creep characteristics were determined at low-temperatures.
" Findings point to improve properties of paving mixes.

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

Article history: The effect of additions of organically modified clay and a sulfur-based compound on the technological, as
Received 29 May 2012 well as rheological properties of a conventional asphalt that was modified by a linear styrene–butadiene–
Received in revised form 26 August 2012 styrene block copolymer, was studied via three factors and five levels of statistical experiment. Blends
Accepted 21 September 2012
were prepared with conventional physical mixing. The linear viscoelastic properties of asphalt binders
Available online 26 October 2012
were studied in small-amplitude sinusoidal oscillations by a dynamic shear rheometer, and the low-
temperature properties of the tested polymer modified asphalts were determined from their creep char-
Keywords:
acteristics measured in a bending beam rheometer. In particular, it was shown that the organoclay had a
PMA
Organoclay
positive influence on rheological properties of the studied systems, specifically on service temperatures.
Rheology Ó 2012 Elsevier Ltd. All rights reserved.
Superpave

1. Introduction The polymer modifiers of asphalt can belong to either the group
of elastomers [1–3] or the group of plastomers [4,5]. Polymer mod-
In recent years, there has been a growing demand for asphalt ified asphalts (PMAs) are usually blends of asphalt with polymer
binders with improved engineering properties, partially in re- added in the range from 2% to 5% by weight. Although there are
sponse to the rising numbers of motor vehicles and axle loads on a variety of polymers, only a few give satisfactory results and can
roads. This all leads to higher stresses on pavements and rapid be used for the preparation of PMAs. The most commonly used
deterioration of pavements (i.e., rutting, cracking). Large seasonal polymers belong to the group of thermoplastic elastomeric block
temperature fluctuations require the use of modified asphalt bind- copolymers; and, from these, styrene–butadiene–styrene (SBS)
ers with wide temperature ranges and good performance proper- especially seems to be a useful modifier of conventional asphalt
ties, particularly in the construction of heavily loaded roads. In [6,7]. Its structure consists of non-compatible blocks that form a
order to prevent rutting and cracking, asphalt binders with addi- two-phase system, where the glassy styrenic domains are dis-
tives of specific properties are frequently produced. One of the persed in the flexible butadienic phase and create a physical elas-
most common methods is the modification of asphalts by various tomeric network. The mixing with asphalt leads to the swelling of
polymers. the SBS by the absorption of maltenes and to the formation of a
spatial network, while the SBS microstructure is maintained [8].
Another possible option for modification is the addition of var-
⇑ Corresponding author at: Department of Civil Engineering, Schulich School of
ious nanoparticles, mostly of inorganic origin, into the polymer
Engineering, University of Calgary, ENF 262, 2500 University Drive NW, Calgary,
Alberta, Canada T2N 1N4. Tel.: +1 403 220 4484; fax: +1 403 282 7026. matrix to enhance the asphalt’s thermomechanical properties, to
E-mail address: martin.jasso@stuba.sk (M. Jasso). improve the barrier properties (in case of waterproofing asphalts)

0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.conbuildmat.2012.09.043
760 M. Jasso et al. / Construction and Building Materials 38 (2013) 759–765

and possibly other properties. The success of the addition of inor- and also polymer and clay. They then prepared a tertiary mixture
ganic particles to polymers depends on the character of the nano- by adding clays and polymer into asphalt, either alone or in the
particles, the organic modifier, the polymeric matrix and the form of a master batch. In both cases, they obtained intercalated
processing technology. nanocomposites with comparable interlayer distances and glass
Layered silicates are commonly used for the preparation of transition temperatures. Their results showed that the mixing
nanocomposites. Polymer nanocomposites (PNCs) can be divided method significantly impacted the final rheological properties of
into intercalated and exfoliated PNCs. Intercalated PNCs are cre- the prepared materials.
ated by the incorporation of polymer chains into the layered sili- The effects of two types of nanoclay (CloisiteÒ and NanofilÒ) on
cate structure that occurs in a crystallographically regular the rheological properties of PMA blends were studied by Jahromi
fashion, regardless of the silicate-to-polymer ratio. Exfoliated PNCs and Khodaii [18]. Asphalt blends were prepared by the melt inter-
are formed by dispersive mixing, where the individual silicate lay- calation process. The dispersion of clay at the nanoscopic level was
ers are separated in a continuous polymer matrix by an average observed with X-ray diffraction (XRD). It was also shown that
distance, which totally depends on the layered silicate loading [9]. Nanofil particles were smaller in size than Cloisite particles. On
Layered silicates with a 2:1 crystallographic structure consist of the other hand, rheological tests have shown clay’s significant
two-dimensional layers, where the central atom of the octahedron effect on the internal structure of asphalt blends, which was
layer of aluminum or magnesium is associated with two external reflected in changes in the rheological properties. The asphalt
tetrahedral layers coordinated by silicon atoms [10]. The layer blends of the studied system showed higher elasticity, lower dissi-
thickness is about 1 nm, and the lateral dimensions of the layers pation of mechanical energy and good compatibility. According to
are in the range of 300 Å to several microns, depending on the type these authors, intercalation of layers of nanoclay depends mainly
of silicates. on asphalt molecules.
For the preparation of PNCs, two properties of the layered sili- Yu et al. [19] studied the effect of nanoclay (montmorillonite)
cates are important. The first is the ability of the silicate particles on the properties of asphalt modified by an SBS copolymer pre-
to disperse at the level of layers, and the second is the ability to pared by melt blending with different amounts of sodium mont-
modify their surface by ion-exchange reactions with organic and morillonite (Na-MMT) and organophilic montmorillonite
inorganic cations. These two characteristics are related to the de- (OMMT). It was found that the addition of Na-MMT and OMMT in-
gree of dispersion of the layered silicate in the polymer matrix creased the viscosity of the SBS-modified asphalt. In addition, the
and depend on the cation in the interlayer. nanoclay/SBS-modified asphalt gained a higher complex modulus
Layered silicates that are miscible with a less polar polymer ma- and lower phase angle, implying a stiffer and more elastic asphalt.
trix should be converted to make an organophilic polymer interca- Therefore, the nanoclay-modified asphalt was determined to have
lation. In general, it is feasible using ion-exchange reactions with better rutting resistance than the original or SBS-modified asphalt.
cationic agents, including primary, secondary, tertiary and quater- Based on the XRD results, it was found that Na-MMT-modified as-
nary alkyl-phosphate or alkyl-ammonium cations. These cations phalt may form an intercalated structure, whereas the OMMT-
reduce the surface energy of the inorganic host and improve the modified asphalt may form an exfoliated structure.
surface wet-ability of polymer aggregates, which results in greater You et al. [20] investigated the influence of larger amounts of
interlayer distances. Moreover, alkyl-phosphate or alkyl- nanoclays as modifiers on the mechanical properties of asphalt
ammonium cations can provide functional groups that can react binders. Nanoclays, at 2% and 4% by weight, were blended with as-
with the polymer matrix and, in some cases, initiate polymeriza- phalt at a high temperature to exfoliate the nanoclay within the as-
tion of the monomers to improve the adhesion between an phalt. It was found that the nanoclays used in this study increased
inorganic compound and a polymer matrix [11,12]. For the prepa- the complex shear modulus (|G|) and viscosity of the prepared
ration of PNCs, four different techniques are used [13–16], namely asphalts binders. The direct tension test also indicated better
in situ polymerization, the template technique, melt intercalation low-temperature cracking resistance. Finally, it is worth noting
and intercalation of polymer from solution. that the blending procedure was crucial in achieving a good distri-
The use of ternary asphalt/polymer/clay system is an unex- bution of nanoclay in asphalt.
plored direction with respect to asphalt modification, and our The addition of two different organoclays as the third compo-
interest was the study of the properties of such individual systems nent of PMAs was tested by Sureshkumar et al. [21]. Mixtures were
prior to the practical design of pavements. For the purpose of as- prepared by adding organoclay and ethylene–vinyl acetate (EVA)
phalt modification, the last two methods (i.e., melt intercalation into asphalt, either individually or as a masterbatch. According to
and intercalation of polymer from solution) seem to be useful. In the chemical structure of the clays used in this work, it was shown
the intercalation of polymer from solution, solvents that are able that the organoclays had a higher affinity to asphalt than to the
to swell silicates and dissolve polymer are used. In this way, lay- polymer; thus, the mechanically mixed clays should preferably
ered silicates with weak interlayer forces are easily dispersed; be intercalated into the asphalt molecules. On the other hand, in
and, polymer chains are able to replace the solvent in the interlayer the case of PMA, there were spotted interactions between the
of silicate. Similarly, it may constitute solvent for polymer, with organoclay and the polymer, which could persist during the mixing
one main difference that asphalt remains in the system as a carrier phase. Master batches of the polymer/clay had higher polarity than
of properties. pure polymers, due to the content of clay. This higher polarity was
In the melt intercalation method, one can either gradually add reflected in improved compatibility with asphalt. Clay also acted as
polymer and clay into a high shear mixer containing the asphalt di- a surfactant. Its presence can lead to increased interaction between
rectly or first prepare a polymer/clay nanocomposite by melt com- the incompatible phases. The authors argued that, generally, the
pounding and then mixing the nanocomposite with asphalt. The higher the affinity between polymer and clay, the better the homo-
latter method is, of course, more time consuming and expensive, geneity of the prepared mixture.
thus less suitable for technological utilization. Liu et al. monitored the impact of organoclay on the properties
Recently, much attention has been devoted to inorganic nanof- of asphalt [22]. Asphalt binders were prepared using a high-shear
illers, which are able (in a relatively small amount) to improve the mixer, with an organoclay content of 4 wt.%. The rheological char-
properties of polymer matrices, as well as asphalt. Polacco et al. acteristics and changes due to the aging of the asphalt were mea-
[17] monitored the effect of adding clays as the third component sured by a dynamic shear rheometer (DSR), and the degree of
to PMAs. These authors studied binary mixtures of asphalt and clay dispersion of organoclay in the asphalt was measured by XRD
 M. Jasso et al. / Construction and Building Materials 38 (2013) 759–765 761

and computed tomography (CT) scans. The presence of organically 2.2.5. Creep tests with a bending beam rheometer (BBR)
modified montmorillonite in the asphalt significantly affected the According to the Strategic Highway Research Program (SHRP), the low-tempera-
ture properties of asphalt binders are determined from their creep characteristics,
rheological properties and resistance to aging, due to the good which are measured in a bending beam rheometer (BBR) according to AASHTO
compatibility between the montmorillonite and the asphalt. The T313. Rectangular beams of asphalt binder (125  12.5  6.25 mm) were cooled in
authors attributed the improved resistance to aging to the organo- a liquid bath, and the deflection was monitored with time and used for calculation
clay particles, which could create an oxygen barrier, thus basically of the stiffness as a function of time. The stiffness (S) and the slope (m) of the stiffness
curve (log S – log t) were then used as the low-temperature characteristics of binders.
preventing the penetration of gas molecules. Another reason may
be the reduced volatility of the oil components of asphalt.
3. Results and discussion
2. Experimental
The influence of variable contents of organoclay on the physical
Conventional asphalt prepared by vacuum distillation of crude oil with a 200/300 and rheological properties and the high-temperature storage sta-
Penetration grade was used as a base for the preparation of polymer/organoclay mod- bility of asphalt binders were studied in [19]. According to these
ified asphalts. This base asphalt was modified by a medium molar mass linear SBS
authors, the storage stability of these modified asphalts decreased
block copolymer with a styrene content of 31 wt.%. The second modifier was Cloisite
CA20 organoclay from Southern Clay Products Inc. The CA20 was prepared from a so- with an increasing content of organoclay. It was found that
dium montmorillonite with a cation exchange capacity of 0.926 meq/g of asphalts modified by organoclay were stable when the content of
Me2(HT)2NCl (dimethyldihydrogenated-tallow ammonium chloride), where hydro- clay was lower than 3 wt.%, there by suggesting the maximum con-
genated tallow is a composition of alkyls on C18 at 65%, C16 at 30% and C14 at 5%. The
tent of organically modified clay in PMAs.
third component was ‘‘Additive A’’, a sulfur-based compound.
In the present study, the influence of changing the concentra-
tion of individual components on the rheological properties of as-
2.1. Preparation of materials
phalt modified by SBS and the above-mentioned organoclays was
The PMAs were prepared using a high-shear mixer. The components of the studied. From a statistical point of view, three factors and five
PMAs were mixed until no separate parts of polymer were observed. The parts of levels of experiments were realized. The three factors were the
the mixtures were gradually dispersed in asphalt at temperatures above 180 °C.
content of the SBS (ranging from 2 to 5 wt.%), the content of Addi-
After mixing, the prepared binders were left to stabilize and were finally stored
in a freezer before further testing. tive A (ranging from 0.01 to 0.12 wt.%) and the content of organo-
clay (ranging from 0.89 to 3 wt.%). These three factors at the five
2.2. Characterization of materials levels of the planned experiment were used for the processing
and statistical evaluation of the measured values [24]. The coded
2.2.1. Penetration values of the experiments are given in Table 2, where the coded
Penetration is an empirical test that provides a general overlook of consistency
levels have been converted into real values for all factors. The com-
of asphalt binders. This test was performed according to the ASTM T5 standard.
positions of PMA blends correspond to levels of the planned exper-
2.2.2. Kinematic viscosity
iment with the introduction of real amounts of SBS, Additive A and
The viscosities of the prepared samples were measured at 135 °C using a Brook- organoclay, see Table 3.
field viscometer DV-III Ultra, equipped with an SC4-31spindle. According to the A detailed analysis of the impact of the individual factors was
AASHTO T316 specification, the maximum viscosity at this temperature must not done via the quality-quantitative evaluation of the statistical signif-
exceed 3000 MPa s.
icance of the effects, as represented by the respective coefficients of
regression [24]. The used regression equations are of the form:
2.2.3. Aging
The effects of paving mix production were studied by using artificial aging of Y ¼ b0 þ b1 x1 þ b2 x2 þ b3 x3 þ b11 x1 x1 þ b12 x1 x2 þ b13 x1 x3 þ b22 x2 x2
the binders in the rolling thin film oven (RTFO), according to AASHTO T240. Aging
in a pressure-aging vessel (PAV) was performed on the binders after the RTFO. The þ b23 x2 x3 þ b33 x3 x3
PAV aging was used for the simulation of long-term asphalt aging in a pavement
during its service life.
where Y is the evaluated parameter; b0, b1, b2, b3, b11, b12, b13, b22,
b23 and b33 are the coefficients of regression; and, x0, x1, x2, x3, x11,
2.2.4. Dynamic rheological properties
x12, x13, x22, x23 and x33 are the factors of the coded levels.
The linear viscoelastic properties of asphalt binders are usually studied in small
amplitude oscillations. These were performed in a DSR, according to AASHTO T315.
The asphalt and mixtures were tested at high and intermediate temperatures and at 3.1. Mechanical properties
the constant frequency of 10 rad/s. The data of the magnitudes of the G and the
associated phase angle (d) were collected in the DSR. The tested binders have to sat-
The aim of these experiments was the acquisition of more exact
isfy the conditions described in the AASHTO T315-09 specification, which are
shown in Table 1. The parameter |G|/sin d estimates the maximum rutting resis- information about a spotted system of PMA binders, which can
tance of the binder, and the parameter |G|sin d is related to fatigue cracking [23]. help in the understanding and, subsequently, setting of the param-
The complex modulus, G(x), was determined in a Malvern Instruments’ CVOR eters of modification, in order to obtain the best properties of PMA
rheometer. The geometry was of the plate–plate type. Original (unaged) and RTFO- binders.
aged samples were studied with plates that were 25 mm in diameter with a gap of
1 mm. PAV-aged samples were tested at intermediate temperatures with 8 mm
The values of the measured characteristics are shown in Table 4.
plates and a gap of 2 mm. The detailed conditions are given in Table 1. The influence of SBS (x1), Additive A (x2) and organoclay (x3) on the
Experimental data were automatically processed by software into the magni- penetration, as measured according to ASTM T5, is shown in Fig. 1.
tude of the |G| and the d, as well as the temperature (the maximum service temper- From this figure, it is evident that only two factors have an effect
ature for the original and RTFO-aged samples, and the intermediate service
on the penetration of PMA mixtures. The most significant impact
temperature for the PAV-aged samples) at which the samples must meet the con-
ditions in Table 1. on the penetration was the SBS, where sharp decreasing values

Table 1
Conditions for high (original and RTFO-aged binders) and intermediate (PAV-aged Table 2
binders) temperatures according to the AASHTO specification. Table of coded values of the planned experiment.

Binder Shear strain (%) Condition Factor (coded level) 1.682 1 0 1 1.682
Original 12 |G|/sin d P 1.0 kPa x1 0.02084 0.02751 0.03729 0.04708 0.05374
RTFO – aged 10 |G|/sin d P 2.2 kPa x2 0.00010 0.00032 0.00065 0.00098 0.00120
PAV – aged 1.0 |G|sin d 6 5000 kPa x3 0.00899 0.01351 0.02014 0.02678 0.03130
762 M. Jasso et al. / Construction and Building Materials 38 (2013) 759–765

Table 3
Conditions for the planned experiment.

Number of Coded levels Real levels SBS content Additive A content Clay content
blends (wt.%) (wt.%) (wt.%)
x1 x2 x3 x1 x2 x3
1 1 1 1 0.02751 0.00032 0.01351 2.64 0.032 1.33
2 1 1 1 0.04708 0.00032 0.01351 4.44 0.032 1.33
3 1 1 1 0.02751 0.00098 0.01351 2.64 0.097 1.33
4 1 1 1 0.04708 0.00098 0.01351 4.44 0.097 1.33
5 1 1 1 0.02751 0.00032 0.02678 2.64 0.032 2.61
6 1 1 1 0.04708 0.00032 0.02678 4.44 0.032 2.61
7 1 1 1 0.02751 0.00098 0.02678 2.64 0.097 2.61
8 1 1 1 0.04708 0.00098 0.02678 4.44 0.097 2.61
9 1.682 0 0 0.02084 0.00065 0.02014 2.00 0.065 1.97
10 1.682 0 0 0.05374 0.00065 0.02014 5.00 0.065 1.97
11 0 1.682 0 0.03729 0.00010 0.02014 3.52 0.010 1.97
12 0 1.682 0 0.03729 0.00120 0.02014 3.52 0.120 1.97
13 0 0 1.682 0.03729 0.00065 0.00899 3.52 0.065 0.90
14 0 0 1.682 0.03729 0.00065 0.03130 3.52 0.065 3.00
15 0 0 0 0.03729 0.00065 0.02014 3.52 0.065 1.97
16 0 0 0 0.03729 0.00065 0.02014 3.52 0.065 1.97
17 0 0 0 0.03729 0.00065 0.02014 3.52 0.065 1.97
18 0 0 0 0.03729 0.00065 0.02014 3.52 0.065 1.97
19 0 0 0 0.03729 0.00065 0.02014 3.52 0.065 1.97
20 0 0 0 0.03729 0.00065 0.02014 3.52 0.065 1.97

of penetration were recorded with an increasing content of SBS in according to AASHTO T316, for all grades, with the limitation that
the mixture. A similar effect was seen also for the organoclay, the maximum viscosity at this temperature must not exceed
where an increasing content of organoclay led to slightly decreas- 3000 MPa s. It can be clearly seen in Fig. 2 that this characteristic
ing values of penetration. The value of penetration of the standard was significantly affected by all factors. Increased loading of the
asphalt was 247 dmm. The presence of all three factors resulted in SBS and organoclay in the asphalt binder led to sharp increases
a decrease of the values of penetration to about 50%. in viscosity, whereas only slightly increased values of viscosity
On the other hand, increasing amounts of Additive A in the mix- can be observed for increased contents of Additive A.
ture caused a slight increase in the value of penetration, but it was There were also interactions between factors, of which the most
not statistically significant. This is clearly seen from Tables 5 and significant were the interactions between the SBS and the organo-
6, which show calculated regression coefficients for mechanical clay, where increasing contents of the SBS and clay shifted the vis-
and rheological properties. The statistically significant coefficients cosity to higher values. The highest measured viscosity of the
are labelled in bold. prepared PMAs was 2206 MPa s, which represents a viscosity that
Testing of asphalts binders according to the Superpave specifi- is ten times higher than standard asphalt (i.e., without modifiers).
cation is the main motivation of this work. It simulates the three
critical stages of a binder’s life. The first stage is represented by 3.2. Rheological properties
the tests carried out on the original asphalts with the goal to
evaluate the abilities of the binder in transportation, storage and For the characterization of the viscoelastic properties of the as-
handling. For this purpose, the viscosity was measured at 135 °C, phalt binders, their dynamic material functions were determined

Table 4
Summary of measured values of mechanical and rheological properties according to the planned experiment.

Number Penetration Viscosity Maximum service Maximum service Intermediate service Low service
of blends @25 °C @135 °C temperature of original temperature of RTFO- temperature of PAV-aged temperature of PAV-
(dmm) (MPa s) mixtures (°C) aged mixtures (°C) mixtures (°C) aged mixtures (°C)
Standarda 247 209 54 52 13 26
1 131 633.4 67.7 65.6 12.1 27
2 107.7 1199.3 76.6 72.3 7.1 23
3 132.7 703.8 69.5 65.4 11.1 27
4 104.7 1443.3 79.8 74.4 9 28
5 126 825.8 71.5 67 10.3 26
6 88 1645.3 78 74.1 7.4 23
7 130 873.8 71.1 71.1 10.6 26
8 96 2206 84.5 76.8 9.7 28
9 140.7 593.9 66.6 62.7 12.1 26
10 92.3 1742 80.1 80.5 5.1 29
11 111.3 958.3 70.7 68.7 12.8 24
12 115.7 1247.3 78.8 70.9 10.1 27
13 125 874.3 73.8 69.4 10.4 27
14 109.7 1360 79.7 71.5 9.3 27
15 119 1042.7 76.5 70.3 11.5 27
16 113.7 1093.7 76.4 70.7 10.5 26
17 115 1080.7 75.3 70.2 10.7 27
18 115.7 1123 75.6 70.7 10.3 27
19 111 1108 76.2 70.2 10.4 27
20 111.3 1111.7 76.4 70.4 9.8 27
a
Standard sample of asphalt without modifiers.
 M. Jasso et al. / Construction and Building Materials 38 (2013) 759–765 763

increasing contents of SBS, Additive A and organoclay in the mix-

ture, the levels of maximum service temperature increased. The
largest share of the increase in temperature was caused by the
maximum content of SBS.
From our statistical evaluation, we can also observe the interac-
tion between Additive A and SBS, which probably led to the forma-
tion of a highly entangled three-dimensional network structure in
PMAs [26] and caused an increase in the values of the monitored
parameter. Another possible explanation maybe that the presence
of organoclay strengthened the network of the internal structure
formed in the PMA. The maximum service temperature of the ori-
ginal asphalt blend was achieved at 84.5 °C.
From the results shown in Fig. 4, one can claim that the maxi-
mum service temperatures of asphalt binders exposed to RTFO
aging were affected by all three factors in the case of the original
Fig. 1. Dependency of the penetration of asphalt mixtures on the levels of the
modified asphalt. The most significant contribution to the growth
planned experiment. of maximum service temperature values was again the amount
of SBS. The impact was four times higher than the impact of
organoclay and Additive A, where the values of the maximum ser-
in a DSR. All tests carried out in the DSR were performed according vice temperature increased with increased content of each factor.
to AASHTO T315. The complex shear modulus (|G|) and its compo- In comparison, in the original and RTFO-aged blends, the maxi-
nents were of particular interest. The |G| is a measure of the total mum service temperature worsened after RTFO-aging was re-
resistance of a material to the deformation when exposed to re- corded, moving in intervals from 2 to 8 °C.
peated pulses of shear stress. The phase angle (d) is an indicator One can suppose that this effect of decreasing maximum service
of the relative amounts of the recoverable and non-recoverable temperatures after RTFO aging of the studied blends was caused by
deformations. Asphalt at high technological temperatures is vis- spotted interactions between the SBS and the organoclay and the
cous and, at low service temperatures, is an elastic material. At SBS and Additive A, i.e., by thermo-oxidative degradation of the or-
intermediate temperatures, asphalt shows true viscoelastic behav- ganic part of the organoclay as well as that of the entire system.
ior with ability to accumulate deformations. This can lead to rut- Another possible explanation of the worsening of this property
ting and further deterioration of pavements. maybe an insufficient dispergation of the organoclay into the
In this regard, the parameter |G|/sin d was suggested. It indi- asphalt, which may lead to agglomeration and, thus, less resistance
cates the rutting potential, especially in conventional asphalt to thermo-oxidative degradation. According to [27,28], the resis-
binders. According to the Superpave specification [25], the asphalt tance to the aging of PMA that have been modified by nanofillers
binder must meet the condition that |G|/sin d is greater than or may be improved, due to the creation of a barrier of intercalated
equal to 1 kPa for the original asphalt binder and greater than and exfoliated structures to oxygen. Based on this work, it can be
or equal to 2.20 kPa for RTFO-aged asphalt binders. For better char- indicated that, with enhancement of the aging resistance, the exfo-
acterization of fatigue cracking at intermediate temperatures, the liated structure is more effective than the intercalated structure.
binders must fulfill the condition of a maximum value of The maximum service temperature of the RTFO-aged asphalt blend
5000 kPa for the parameter |G|sin d. In this case, low values of both was achieved at a value of 80.5 °C.
|G| and d are considered desirable for better fatigue cracking After RTFO aging, samples were exposed to simulated long-term
resistance. aging (estimated 5–10 years of service life) through the PAV. These
In Fig. 3, the influences of the contents of SBS, Additive A and samples were tested at intermediate service temperatures (i.e.,
organoclay on the maximum service temperature, which were ob- where the intermediate service temperature meets the condition
tained from (|G|/sin d), can be observed as the last characteristic of of |G|sin d 6 5000 kPa), as well as at low service temperatures.
the original modified asphalt. From the results, we can conclude The intermediate service temperature characteristic is portrayed
that the maximum service temperature of the original asphalt in Fig. 5, where we can see that, with increased content of SBS in
binders was again significantly affected by all three factors: with the mixture, the intermediate service temperatures at which the

Table 5
Regression coefficients of mechanical and rheological properties of modified asphalts.

Parameter Penetration @25 °C Viscosity @135 °C Maximum service Maximum service Intermediate service Low service
(dmm) (MPa s) temperature of original temperature of RTFO- temperature of PAV-aged temperature of PAV-
mixtures (°C) aged mixtures (°C) mixtures (°C) aged mixtures (°C)
bi bic bi bic bi bic bi bic bi bic bi bic
b0 114.328 3.138 1090.045 30.225 75.982 0.521 70.410 0.243 10.542 0.588 -26.82 0.428
b1 14.992 2.082 394.696 20.058 4.527 0.346 4.280 0.161 1.660 0.390 0.077 0.284
b2 1.326 103.226 1.811 0.908 0.076 1.102
b3 4.528 174.929 1.569 1.086 0.231 0.147
b12 0.087 2.720 85.788 26.202 1.038 0.451 0.113 0.211 0.613 0.510 1.250 0.371
b13 2.588 105.788 0.088 0.363 0.413 0.250
b23 1.663 36.788 0.137 0.612 0.213 0.000
b11 0.496 2.028 39.708 19.536 0.929 0.337 0.463 0.157 0.741 0.380 0.112 0.277
b22 0.565 16.625 0.433 0.174 0.267 0.596
b33 0.797 21.710 0.275 0.056 0.299 0.065

Note: Coefficients labelled in bold are statistically significant at a significance level of 95%.
764 M. Jasso et al. / Construction and Building Materials 38 (2013) 759–765

Table 6
Statistical parameters.

Statistical Penetration Viscosity Maximum service Maximum service Intermediate service Low service temperature
parameter @25 °C @135 °C temperature of original temperature of RTFO-aged temperature of PAV-aged of PAV-aged mixtures
(dmm) (MPa s) mixtures (°C) mixtures (°C) mixtures (°C) (°C)
FS1 125.48 1078.72 484.09 1722.9 40.72 33.91
FS2 1.75 37.76 17.07 24.70 8.09 18.55
SLF 0.42 15.03 8.97 53.89 4.60 10.335
SE 1 1 1 1 1 1

Note: FS1 – Fisher–Snedecor criteria for confidence testing of the linear part of the regression equation; FS2 – Fisher–Snedecor criteria for confidence testing of the nonlinear
part of the regression equation; SLF – standard deviation of the lack of fit error source; SE – standard deviation of experimental error source.

Fig. 2. Dependency of the viscosity of asphalt mixtures on the levels of the planned Fig. 4. Dependency of the maximum service temperature of the RTFO-aged asphalt
experiment. mixtures on the levels of the planned experiment.

Fig. 5. Dependency of the intermediate service temperature of the PAV-aged

Fig. 3. Dependency of the maximum service temperature of the original asphalt
asphalt mixtures on the levels of the planned experiment.
mixtures on the levels of the planned experiment.

3.3. Low-temperature cracking

tested samples met the Superpave specification requirement (|G|-
sin d 6 5000 kPa) were linearly decreasing with an increasing con- Thermal cracking is observed at low temperatures, when an as-
tent of Additive A and simultaneously passing through a maximum phalt binder shows a tendency to shrink, leading to the increase of
depending on the content of SBS. tensile stresses in pavements. The low-temperature performance
The effect of individual Additive A and individual organoclay was tested according to AASHTO T313. The measured values,
additions have been proven as statistically irrelevant; however, it which are shown in Fig. 6, were also investigated by an analysis
is worth mentioning that an interaction between Additive A and of variance (ANOVA). One can establish that only Additive A statis-
the SBS was statistically significant, and that an increasing content tically affected the low service temperature. Increasing the content
of Additive A led to lower values of this temperature. The lowest of Additive A led to decreasing values for the low service temper-
value of intermediate service temperature parameter, according ature, up to a concentration of 0.065 wt.%. Another addition of
to planned experiment, was achieved at the value of 5.1 °C, while Additive A caused increasing values of this property.
the value of the intermediate service temperature parameter for It is worth mentioning that there were interactions between
the standard asphalt sample was 13 °C, indicating an improvement factors x1 and x2, i.e., SBS and Additive A. While the increase
of this parameter by 8 °C. of SBS caused worsening of the low service temperature, its
 M. Jasso et al. / Construction and Building Materials 38 (2013) 759–765 765

These findings point to possibility of using quaternary PMAs for

the preparation of paving mixes; however, additional research in
this area is necessary.

Acknowledgements

The authors would like to acknowledge Husky Energy Inc. for its
financial support of this work. We would like to thank Mr. R. Wirth
and Mrs. E. Vargova, for their help with the preparation of the
PMAs, and Dr. J. Stastna, for valuable discussions on the rheological
properties of PMAs.

References

[1] Yildirim Y. Polymer modified asphalt binders. Constr Build Mater

2007;21(1):66–72.
[2] Vlachovicova Z, Wekumbura C, Stastna J, Zanzotto L. Creep characteristics of
Fig. 6. Dependency of the low service temperature of the PAV-aged asphalt asphalt modified by radial styrene–butadiene–styrene copolymer. Constr Build
mixtures on the levels of the planned experiment. Mater 2007;21(3):567–77.
[3] Xiang L, Cheng J, Que G. Microstructure and performance of crumb rubber
modified asphalt. Constr Build Mater 2009;23(12):3586–90.
[4] Panda M, Mazumdar M. Utilization of reclaimed polyethylene in bituminous
interaction with Additive A led to a sharp shift of this property to paving mixes. J Mater Civ Eng 2002:527–30.
lower (i.e., better) values. Experimental values ranged from 23 [5] Giavarini C, De Filippis P, Santarelli ML, Scarsella M. Production of stable
polypropylene-modified bitumens. Fuel 1996;756:681–6.
to 29 °C, which represents a temperature difference of 3 °C in
[6] Airey GD. Rheological properties of styrene butadiene styrene polymer
comparison with the standard sample. modified road bitumens. Fuel 2003;82:1709–19.
[7] Airey GD. Styrene butadiene styrene polymer modification of road bitumens. J
Mater Sci 2004;39:951–9.
[8] Vonk W, Hartemink R. SBS-modified binders, also cost effective in hot
climates; 2004. [ISBN 1-920-01718-6].
4. Conclusions [9] Ray SS, Cele N. Recent progress on nafion-based nanocomposite membranes
for fuel cell applications. Macromol Mater Eng 2009;294:719–38.
[10] Ray SS, Okamoto M. Polymer/layered silicate nanocomposites: a review from
The impact of adding organically modified clay and Additive A preparation to processing. Progr Polym Sci 2003;28:1539–641.
on the technological, as well as rheological, properties of a conven- [11] Blumstein A. Polymerization of adsorbed monolayers: II Thermal degradation
of the inserted polymers. J Polym Sci A 1965;3(1):2665–73.
tional asphalt modified by a linear SBS block copolymer were stud-
[12] Messersmith PB, Giannelis EP. Synthesis and barrier properties of poly(e-
ied via three factors and five levels of statistical experiments. From caprolactone)-layered silicate nanocomposites. J Polym Sci, Part A: Polym
the experimental point of view, the method was used to acquire Chem 1995;33:1047–57.
[13] Shin S-YA, Simon LC, Soares JBP, Scholz G. Polyethylene-clay hybrid
useful and comprehensive information. The blends were prepared
nanocomposites: in situ polymerization using bifunctional organic modifiers.
by conventional physical mixing, and the mechanical and rheolog- Polymer 2003;44(18):5317–21.
ical properties were studied. [14] Zhong Y, Zhiyong Z, Wang S-Q. Synthesis and rheological properties of
The viscosity of prepared blends increased and the penetration polystyrene/layered silicate nanocomposite. Polymer 2005;46:3006–13.
[15] Okada A, Fukushima A, Kawasumi M, Inagaki S, Usuki A, Sugiyama S, Kurauchi
decreased, depending on the increasing content of the components T, Kamigaito O. Composite material and its preparation. US Patent 4739007;
in the asphalt blends. 1987.
The maximum service temperature of the original asphalt bind- [16] Jordan JW. Organophilic bentonites. I. Swelling in organic liquids. J Phyl Col
Chem 1949;53(1):294–306.
ers was affected by all three factors. With increased content of [17] Polacco G, Filippi S, Stastna J, Biondi D, Zanzotto L. Rheological properties of
individual components in the mixtures, there were increases in asphalt/SBS/clay blends. Eur Polym J 2008;44:3512–21.
the maximum service temperatures. One can claim that improve- [18] Jahromi SG, Khodaii A. Effects of nanoclay on rheological properties of bitumen
binder. Constr Build Mater 2009;23(8):2894–904.
ment of the maximum service temperature can result from the for- [19] Yu J, Zeng X, Wu S, Wang L, Liu G. Preparation and properties of
mation of a highly entangled three-dimensional network structure montmorillonite modified asphalts. Mater Sci Eng A 2007;447(1–2):233–8.
in polymer modified asphalts. [20] You Z, Mills-Beale J, Foley JM, Roy S, Odegard GM, Dai Q, et al. Nanoclay-
modified asphalt materials: preparation and characterization. Constr Build
The worsening of the maximum service temperatures of the as-
Mater 2011;25(2):1072–8.
phalt binders that were exposed to RTFO aging was also monitored. [21] Sureshkumar MS, Filippi S, Polacco G, Kazatchkov I, Stastna J, Zanzotto L.
This worsening or decreasing of the maximum service tempera- Internal structure and linear viscoelastic properties of EVA/asphalt
nanocomposites. Eur Polym J 2010;46(4):621–33.
tures was caused by spotted interactions between the SBS and
[22] Liu G, Wu S, van de Van M, Yu J, Molenaar A. Influence of sodium and organo –
the organoclay and the SBS and Additive A, i.e., by thermo- montmorillonites on the properties of bitumen. Appl Clay Sci 2010;49(1–
oxidation degradation or insufficient dispergation of organoclays 2):69–73.
into asphalt binders. [23] Navarro González M. Rheology and engineering parameters of bitumen
modified with polyolefins, elastomers and reactive polymers. Technische
The |G|sin d of less than or equal to 5000 kPa at intermediate Uni Berlin; 2010.
service temperatures were statistically affected by Additive A and [24] Montgomery DC. Design and analysis of experiments. 5th ed. New York: John
SBS. Increasing the content of these components and their interac- Wiley & Sons; 2001.
[25] American Association of State Highway and Transportation Officials; 2009.
tions led to lower values of the fatigue-cracking temperature. [26] Shin EE, Bhurke A, Scott E, Rozeveld S, Drzal LT. Microstructure, morphology,
Similarly, in the case of low service temperatures, increasing and failure modes of polymer-modified asphalt. Transport Res Board Rec
contents of Additive A led to decreasing values for the low service 1996;1535:61–73.
[27] Yu J, Wing L, Zeng X, Wu S, Li B. Effect of montmorillonite on properties of
temperature. In spite of the fact that increased content of SBS styrene-butadiene-styrene copolymer modified bitumen. Polym Eng Sci
caused worsening of the low service temperature, interaction of 2007;47(9):1289–95.
the SBS with Additive A led to a sharp shift of this property to lower [28] Yu J-Y, Feng P-Ch, Zhang H-L, Wu S-P. Effect of organo-montmorillonite on
aging properties of asphalt. Constr Build Mater 2009;23(7):2636–40.
values.