-
Rheological, microscopic, and chemical characterizationof the
rejuvenating effect
Xiaokong Yu a,, Martins ZaumanaWorcester Polytechnic Institute,
100 Institute Road, USb EMPA Swiss Federal Laboratories for
Materials Science
h i g h l i g h t s
The rejuvenating effect for aged binders was tal comfect ben
o commonly used rejuvenators. The rheological properties of the
virgin, aged,ere tested using the dynamic shear rheometer and the
bending beam rheom-
All rights reserved.
1. Introduction
As of 2011, more than 40 state agencies in the US allow to
useabove 30% reclaimed asphalt pavement (RAP) in mix design of
Hot
Mix Asphalt (HMA) [1], however, currently the average RAPstill
between 10% and 20% [2]. According to the Superpadesign specication
(AASHTO M 323), when more than 15% RAPis used in the mix design, it
is required to reduce the binder perfor-mance grade (PG) by one to
compensate for the aged RAP binder.Furthermore, when adding more
than 25% RAP, virgin binder PGhas to be determined based on the
properties of extracted RAP bin-der. Such requirements are set to
ensure that the virgin bindercompensates for the deteriorated
mechanical properties of the
Corresponding authors. Tel.: +1 508 831 5530; fax: +1 508 831
5808 (X. Yu).Tel.: +41 58 765 4801; fax: +41 58 765 1122 (S. dos
Santos).
E-mail addresses: [email protected] (X. Yu),
[email protected](S. dos Santos).
Fuel 135 (2014) 162171
Contents lists availab
Fue
.e 2014 Elsevier
Ltd.http://dx.doi.org/10.1016/j.fuel.2014.06.0380016-2361/ 2014
Elsevier Ltd. All rights reserved.use isve mixSARA
fractionationMechanical properties eter. Furthermore, in order to
better understand the rejuvenating effect, surface microscopic
properties
and chemical composition of the binders were measured using
atomic force microscopy (AFM) and SARA(Saturates, Aromatics,
Resins, Asphaltenes) fractionation, respectively. Results indicated
that the bulkmechanical properties (complex modulus and viscosity)
of the rejuvenated binders were in betweenthose of the virgin and
aged binders. Aging and rejuvenation led to morphological changes
as comparedto their virgin binders; however, the rejuvenated
binders did not always reproduce the microstructures ofthe virgin
binders. Microscopic measurements on adhesion and dissipation of
virgin, aged, and rejuve-nated samples were qualitatively
consistent with the bulk rheological results. SARA separation
resultssuggested that changes in chemical fractions were
responsible for the stiffening effect of aging and theimprovement
of mechanical properties with the addition of the rejuvenators.
Such a systematic approachof characterizing the rejuvenating
mechanism will benet the effort of producing more sustainable
RAP-containing asphalt pavements.RejuvenatorAFM complete blending
with tw
and rejuvenated binders w Results suggest that changes in chemic
Characterization of the rejuvenating ef
a r t i c l e i n f o
Article history:Received 4 April 2014Received in revised form 9
June 2014Accepted 12 June 2014Available online 2 July 2014
Keywords:Asphalt binderon asphalt binders
is a, Salom dos Santos b,, Lily D. Poulikakos bAand Technology,
berlandstrasse 129, 8600 Dbendorf, Switzerland
horoughly evaluated at both macro- and micro-scales.positions
contributed to mechanical properties variations.ets the effort for
better recycling bituminous materials.
a b s t r a c t
With the increasing environmental awareness and rising costs of
virgin binders, reclaimed asphalt pave-ment (RAP) has been used as
an alternative for energy and cost saving in asphalt pavements.
However,RAP binders have been aged to different extents during
pavements service life and adding rejuvenatingagents provides a
practical means for restoring the mechanical properties of the aged
binders reducingthe needed additional virgin binder. In many
studies, the rejuvenating effect has been evaluated in termsof the
improvement of rejuvenated binders rheological properties whereas
the fundamental rejuvena-tion mechanism remains unclear. In this
research, two different asphalt binders from the Materials
Ref-erence Library of the Strategic Highway Research Program (SHRP)
were aged, and rejuvenated byjournal homepage: wwwle at
ScienceDirect
l
lsevier .com/locate / fuel
-
ing the mechanical properties of aged binders) is a
relativelypoorly understood alternative for mix design of
RAP-containing
2.1.2. RejuvenatorsTwo generic rejuvenators, aromatic extract
and waste vegetable
oil (WV oil), were used in the study. Table 3 shows their
viscosityand specic gravity measured by the authors and other
relevantproperties obtained from the manufacturers. A brief
descriptionof the two rejuvenators is also provided below. Based on
previousstudies [8,9], 12 wt.% of each rejuvenator was added to the
binders.
Aromatic extract is a traditional rejuvenator with dominantpolar
aromatic molecules. According to the manufacturer, aromaticextract
contains approximately 75% of aromatic oil and resin com-pounds
with small amount of saturate oil. Chemical compositionsof the
aromatic extract were determined in this study using
SARAfractionation and discussed later (Section 3.3).
WV oil is increasingly used for bio-diesel production. It
isderived from fast and convenience food frying oil and also
referredto as yellow grease. This product usually has low free
fatty acidcontent (
-
eleumase
135treatment (i.e., rejuvenation) before application for another
serviceperiod.
2.2.2. Addition of rejuventorsThe aged binders were dosed in
containers at room tempera-
ture, thereafter heated at 145 C for 40 min and blended for5 min
with 12 wt.% of each rejuvenator by hand. The samples
wereimmediately re-heated for 10 min to remove the air
bubbles,blended again for another 2 min, and thereafter left to
cool at roomtemperature. Each blend was then re-heated at 140 C for
30 minbefore sampling into containers suited for each test.
2.2.3. Dynamic shear rheometer (DSR) measurementsThe rheological
properties of virgin, aged, and rejuvenated
asphalt binders were measured using a DSR (Anton Paar PhysicaMCR
301). According to AASHTO T 315, an 8 mm plate-plate geom-etry with
2 mm gap in strain controlled mode was used. The testswere
performed at 40, 50, 60, 70, and 80 C and frequencies of0.120 Hz at
each temperature. These results were shifted to areference
temperature of 40 C using the timetemperaturesuperposition
principle to construct master curves of complexmodulus (G) and
phase angle (d). Shift factors were computedusing
WilliamsLandelFerry (WLF) Eq. (1) and a sigmoidalfunction dened by
Eq. (2) was used for tting the shifted shearmodulus data. Larger
values of G correspond to stiffer asphaltbinders, and larger values
of d correspond to larger viscous compo-nent of the complex
modulus. Interpretation of master curvesallows to describe the
temperature dependence of the viscoelasticbehavior of the
binder.
log aT c1T TrC2 T Tr 1
log G d a1 ebclogaTlogg 2
where,
aT horizontal shift factor at temperature TT test temperature,
CTr reference temperature, C (40 C in this study)f testing
frequency, rad/sd, a, b, c, C1, C2 tting parameters determined
using leastsquares method.
2.2.4. Bending beam rheometer (BBR) measurementsBBR tests
(AASHTO T 313) were performed for all samples in
order to characterize the low temperature cracking potential
(i.e.,the creep stiffness and creep rate). To increase the
resistance tothermal cracking, a lower stiffness is favorable to
lessen the ther-
Table 3Properties of the rejuvenators [10].
Rejuvenator Viscosity at 135 C (poise) Specic gravity Sourc
Aromatic extract 9.20e4 0.995 PetroWaste vegetable oil 5.17e4
0.917 Bio-b
164 X. Yu et al. / Fuelmal stresses developed in the pavements
and a high creep rate(also called m-value) is desirable to rapidly
disperse the accumu-lated stress. Superpave specications require a
maximum stiffnessof 300 MPa and a minimum m-value of 0.3 at the
specied testtemperature. The minimum pavement design temperature
isdened to be 10 C lower than the test temperature. In this study,a
single test temperature of 18 C was chosen to allow
directcomparison of all samples under the same conditions. This
temper-ature would correspond to Superpave low PG of the AAD
binder(28 C) but is much lower than the PG of the ABD binder(10 C).
For each sample, the average values of creep stiffnessand creep
rate from two repetitive tests are reported.
2.2.5. Atomic force microscopy (AFM) measurementsWith a sharp
tip attached at the free end of a cantilever probing
the specimen surface, AFM is able to measure the materials
sur-face morphology and mechanical properties with high
resolution.Through microscopic investigations on how the
rejuvenator inter-acts and modies the aged binders at the micro
scale, AFM charac-terization can provide complementary information
on therejuvenation mechanism of aged binders.
Thin lms of asphalt binders were prepared by heat-castmethod,
which maintains binders solid-state microstructure[12]. For
morphological measurements, a bead of binder (ca.50 mg) was dropped
onto a glass slide, which was heated for2 min in an oven at 115 C.
Once the binder became liquid, itwas spread out with a blade
(buttering) to form a thin lm. Thishot lm was left undisturbed for
an additional 10 min to allowthe surface to ow to a smooth nish.
For surface mechanical prop-erty measurements that were conducted
in a different laboratoryusing a different AFM, bitumen (ca. 10 mg)
were spread over a0.9 0.9 mm2 area of a glass slide and the samples
were placedin the oven at 110 2 C for 20 min. The lm was then
coveredto prevent dust pick-up and left to cool in air to room
temperature(20 C), and stored for a minimum of 24 h before
imaging.
Asphlat binders morphology was characterized using an Asy-lum
Research AFM (MFP-3D) with intermittent-contact mode asit minimizes
the possible tip contamination caused by bindersadhesiveness [13].
Budget Sensors silicon tips Tap300Al-G (nomi-nal resonant frequency
of 300 kHz and nominal stiffness of 40 N/m) were used for
morphological measurements. Tip geometryand stiffness were
calibrated using a reference sample (TGT 01,MicroMasch) and a
thermal tune procedure, respectively. Asphaltbinders mechanical
properties were measured using a BrukerAFM (MultiMode 8) with a
Nanoscope V controller in peak forcetapping quantitative
nanomechanical (PFT QNM) mode. The datawas collected using
Nanoscope 8.15. Bruker Tap150A tips (nominalresonant frequency of
150 kHz and nominal stiffness of 5 N/m)were used. The tip radius
was calibrated using the Bruker PS-LDPEreference and set as 20 nm.
The peak force was set as 15 nN, thefeedback gain was 20, and the
scan rate was 0.5 Hz. The peak forceamplitude and peak force
frequency were set as 150 nm and 2 kHz,respectively. Scans of the
topographical and mechanical propertieswere recorded over 40 40 lm2
or 20 20 lm2 area with256 256 pixels, depending on the sizes of
their specic micro-structures. All measurements were conducted at
room tempera-ture (20 C).
PFT QNM mode is a technique recently developed by Bruker,and it
allows simultaneous capture of topographical and mechan-
Rened or waste Dominant molecular structure Relative
polarity
Rened Aromatic Polard Waste Ring and stranded Non-polar
(2014) 162171ical properties by recording instantaneous force
curves as the AFMprobe approaches and rectratcs from the sample
surface as shownin Fig. 1. As the AFM tip approaches the surface,
it experiences anattractive force which causes the tip to jump into
contact with thesample. After contact, repulsive forces dominate
the tip-sampleinteraction, leading to a peak force point in the
approaching curve.As the tip is retracted, it goes through a
minimum force corre-sponding to the adhesion force and, nally, the
contact breaksapart. In PFT QNM mode, the peak force and the
indentation depthcan be controlled. The surface mechanical
properties of asphalt
-
ASTM D4124-09 2010. Asphaltenes and maltenes were separated
v/v) and 150 ml trichloroethylene (ACS grade). Saturates were
col-
135binders (i.e., adhesion, dissipation energy, and elastic
modulus) areextracted from force curves that are acquired at each
pixel of animage and, therefore, a map of each mechanical property
isobtained. The adhesion force, also called pull-off force,
correspondsto the lowest point of the retraction curve. Dissipation
energy isdened as the area in between the approach and retraction
curves,which is related to a materials damping property. To obtain
amaterials elastic modulus, a portion of the retraction curve
(i.e.,3090% from the adhesion force to the maximum peak force,
cor-responding to the linear portion of the curve (Fig. 1)) can be
ttedusing different contact mechanics models (e.g., Hertz [14],
Derja-guinMllerToporov (DMT [15]), JKR (JohnsonKendallRoberts[16],
and Maugis [17]). Among these contact mechanics, the Hertzmodel
does not consider adhesion; DMT is appropriate for stiffmaterials
with low adhesion; JKR provides good prediction forlarge tip radii
and compliant samples with high surface forces.The Maugis mechanics
proposes a parameter lambda, k, denedin Eq. (3), to characterize a
wide spectrum of materials: largelambda (k?1) is for the more
compliant, adhesive combinations,approaching JKR mechanics, and for
small lambda (k? 0), the tip-sample system responds like the DMT
mechanics.
k 2:06n
R x22
1=33
Fig. 1. Schematic force curves from a typical AFM PFT QNM
measurement duringapproach and retraction of the tip.
X. Yu et al. / Fuel0 pE
where n0 is the interatomic spacing of the tip-sample system, R
isthe effective radius of curvature of the tip-sample system, x
isthe work of adhesion, and E is the reduced elastic modulus,
dened
as E 1v2tEt 1v2sEs
h i1, in which v is the Poissons ratio, and the sub-
scripts t and s refer to the tip and sample, respectively. For
the tip-binder system, by using the relative parameters from the
availableliterature (n0, R, x and K to be 3 , 10 nm, 60 mJ/m2, and
890 MPa,respectively), the value for k is calculated to be about
10, largeenough indicating that JKR might be more suitable for
capturingthe tip-binder contact mechanics. However, PFT QNM adopts
theDMT model rather than the JKR for real-time modulus mapping ofa
material, which simplies the calculation of modulus but maylead to
discrepancies that should be noted. The DMT contactmechanics model
is shown in Eq. (4),
F Fadhesion 43 E
Rd3
p4
where F is the instantaneous force on the tip, Fadhesion is the
adhe-sion force, E and R are the same as in Eq. (3), and d is the
tip-sampleindentation depth.lected prior to elution of a uorescent
band migrating up the col-umn. Aromatics characterized by the
uorescent band werecollected prior to a dark band migrating up the
column belowthe uorescent band. Trichloroethylene was nally
introduced tostrip the column of any remaining dark band material
denedas the polar aromatics fraction. Fractions were recovered by
roto-evaporation and dried to a constant weight.
2.2.7. Test planThe testing matrix is shown in Table 4. WV oil
rejuvenated AAD
sample was too soft to complete DSR successfully in the
tempera-ture range of 4080 C as well as the BBR test at 18 C. SARA
frac-tionation was performed on the virgin, aged, and aromatic
extractrejuvenated samples for both binders and also the aromatic
extractalone. Microscopic morphologies of virgin, aged, and
rejuvenatedbinders were obtained successfully with
intermittent-contactmode AFM. Furthermore, preliminary measurements
of surfacemechanical properties on AAD-based samples faced
technical chal-lenges such as tip contamination, excessive
deformation under arelatively small compressive force and
complicated tip-sampleinteraction due to the dominant
bee-structures on the samplesurfaces (as shown in Fig. 3). WV oil
rejuvenated ABD samplewas too compliant under a compressive force
that was used forother ABD-based binders. Therefore, AFM mechanical
measure-ments were only carried out for virgin, aged, and aromatic
extractrejuvenated ABD samples.
3. Results and discussion
3.1. Rheology
3.1.1. Properties at intermediate and high temperaturesThe
rheological properties of the virgin, aged, and rejuvenated
asphalt binders at intermediate and high temperatures
obtainedusing isooctane (2,2,4-trimethyl pentane, HPLC grade,
Fisher Scien-tic, Waltham, MA, USA). The precipitating solvent was
added tothe binder sample at a ratio of 40:1 by mass, and the
mixturewas stirred overnight at room temperature. The sample
solutionwas then ltered through both a Fischer brand lter paper Q2
(poresize 15 lm) and a Millipore cellulose ester membrane (pore
size0.22 lm). The asphaltenes retained on the lter paper werewashed
with additional solvent until the ltrate was colorlessand dried in
a fume hood until a constant weight was obtained.The maltene
fraction was recovered by roto-evaporation and driedto a constant
weight. Separation of the maltene into saturates, aro-matics and
resins was carried out by injecting the solution (1 g ofthe maltene
dissolved into 10 ml of n-heptane) into a 70-cm long,105-cm
diameter glass LC (liquid chromatography) column withapproximately
100 g of chromatographic grade activated alumina(80200 mesh, VMR
International, US). The three fractions wereeluted by introducing
the following solvents (all by Fisher Scien-tic) into the column:
50 ml n-heptane (HPLC grade), 100 ml of tol-uene (HPLC grade), 75
ml of methanol (HPLC grade)/toluene (50:502.2.6. SARA
fractionationSARA (saturates, aromatics, resins, asphaltenes)
fractionation is
a frequently used technique for chemical compositional analysis
ofasphalt binders from different crude sources and binders
subjectedto various treatments [18]. Resins are sometimes referred
as polararomatics while aromatics as naphthenic aromatics [19]. In
thisstudy, SARA fractionation was conducted according to the
standard
(2014) 162171 165from DSR are shown in Fig. 2.Virgin ABD was
stiffer (higher complex modulus) yet more vis-
cous (larger phase angle) than virgin AAD as shown in the
master
-
Agex
135Table 4Testing matrix.
Test method Analysis Aromaticextract
ABD binder
Virgin Aged
DSR Rheology at intermediate &high temperatures
BBR Rheology at low temperature AFM
intermittent-Morphology
166 X. Yu et al. / Fuelcurves, which supports the results in the
SHRP report (Table 2) [7].For both binders, aging resulted in an
increase of the complex mod-ulus and a decrease of the phase angle
(corresponding to a lowercomplex viscosity), which conforms with
the well-known agingeffect [20,21]. The addition of the
rejuvenators into the aged sam-ples decreased the complex modulus
and increased the phaseangle in comparison to aged samples, to
different extents depend-ing on both the binders crude sources and
the rejuvenating agents.
Comparing the effect of the two rejuvenators on one type
ofbinder (ABD), adding 12 wt.% of aromatic extract into the agedABD
binder restored the complex modulus to the level of the virginABD
and the phase angle in between the virgin and aged samples.
contactAFM PFT QNM Surface mechanical properties SARA
fractionationChemical composition
Fig. 2. Complex modulus (a) and phase angle (b) master curves of
virgin, aged, andAAD binder
ed + Aromatictract
Aged + WVoil
Virgin Aged Aged + Aromaticextract
Aged + WVoil
(2014) 162171Whereas the same dosage of WV oil reduced the
complex modulusbelow the virgin level and increased the phase angle
above the vir-gin level, which are not necessarily desirable
features. In thisrespect, 12 wt.% of aromatic extract achieved the
desired level ofrejuvenation on this binder compared to the same
dosage of theWV oil. These results show that in the case of
complete blending,the restoration of rheological properties can be
achieved usingrejuvenators, the extent of which is binder source
and rejuvenatordependent.
The rejuvenating efciency of 12 wt.% of aromatic extract on
thetwo binders (AAD and ABD) in terms of restoring their
rheologicalproperties was similar. For both binders, the complex
modulus
rejuvenated asphalt binders AAD and ABD at a reference
temperature of 40 C.
-
135X. Yu et al. / Fuelmaster curves for the rejuvenated binders
almost overlapped withtheir source virgin binders. Values of the
phase angle for therejuvenated binders fell in between the virgin
and aged samples.
Fig. 3. Topographic images of virgin (top row), aged (2nd row),
aromatic extract (3rd40 40 lm2) measured at room temperature (20
C). The color scales range over 10 n(2014) 162171 167However, aging
and rejuvenation caused more signicant shift ofphase angle for
AAD-based samples than ABD-based binders atlower frequencies
(corresponding to higher temperature). This
row), and WV oil rejuvenated (4th row) ABD (left, 20 20 lm2) and
AAD (right,m and 60 nm for ABD and AAD based samples,
respectively.
-
can be attributed to the rheological differences of the two
bindersfrom different crude sources.
3.1.2. Properties at low temperature
Furthermore, as noted by Allen et al. [24] the complicated
molecu-lar interactions among the various chemical components
manifeststhemselves into the formation of the microstructures.
The effect of aging on the two binders can be seen in the
secondrow in Fig. 3. Upon aging, the size of the ake-like
microstructures
the bee-structures and the matrix. The introduction of the WV
oilinto the aged binders (both AAD and ABD) did not modify
their
BD.
t
168 X. Yu et al. / Fuel 135 (2014) 162171The BBR results of the
virgin, aged, and rejuvenated asphaltbinders at 18 C are reported
in Table 5. In accordance with theirPG grades, the virgin ABD
binder (PG 10 C) is much stiffer andhas a lower m-value compared to
AAD (PG 28 C). As expected,aging increased stiffness while reducing
m-value for both binders;the introduction of the rejuvenators
reversed this trend to a certaindegree.
Considering the ABD binder, consistent with the observations
atintermediate and high temperatures discussed in the previous
sec-tion, WV oil was more effective in softening the binder
comparedto aromatic extract. It produced a binder that was
signicantlysofter than the virgin ABD and in fact would pass the
Superpaverequirement for PG 28 C (Superpave grade is 10 C lower
thanthe test temperature). Aromatic extract softened the aged ABD
bin-der to a level below that of the original virgin ABD, while its
reju-venation effect on aged AAD binder was different with a
stiffnessslightly higher and an m-value slightly lower than those
of the vir-gin binder. This is somewhat different from the
observations atintermediate and high temperatures where the effect
of aromaticextract was qualitatively similar for both types of
binders. Thedependence of the rejuvenating effect of aromatic
extract on bin-der type and test temperature is probably related to
the chemicalvariations between their crude sources.
3.2. Microscopic morphology and mechanical properties using
AFM
DSR and BBR tests indicated that when completely blended,
therejuvenators were capable of restoring the rheological
propertiesof the aged binders close to the level of their source
virgin binders.In order to better understand the rejuvenating
mechanism, surfacemicroscopic morphology and mechanical properties
characteriza-tion using AFM was carried out.
3.2.1. Microscopic morphologyMicrostructures of the virgin,
aged, and rejuvenated AAD and
ABD samples are shown in Fig. 3. Scan sizes for AAD- and
ABD-based samples are different and related to the sizes of their
charac-teristic microstructures. The grayscales for each binder
were set tobe the same (60 nm and 10 nm for AAD and ABD,
respectively) foreasy visual comparison among the samples under
differenttreatments.
As shown in the top row in Fig. 3, the two virgin binders
dis-played considerably different morphologies, even though both
ofthem contain dispersed and continuous domains. Virgin ABD hasa
dispersed phase with ake-like structures (with an average sizeof
less than 2 lm in diameter) spreading over a smooth matrixphase;
whereas elliptical domains containing the bee-structures(with major
andminor axes of a fewmicrons) dominate the surfaceof the virgin
AAD. The appearance of the bee-structures hasrecently been
attributed to the interaction between crystallizingwaxes and the
remaining non-wax asphalt components [22,23].The morphological
difference between the two binders can berelated to their chemical
compositional differences as shown inTable 1, most notably the
higher wax content of the AAD binder.
Table 5BBR results at 18 C for the virgin, aged, and rejuvenated
asphalt binders AAD and A
Parameter ABD binder
Virgin Aged Aged + Aromatic extracCreep stiffness (MPa) 760 839
578m-value 0.205 0.159 0.274morphologies signicantly, as compared
to the aged samples (thefourth row in Fig. 3).
In summary, the addition of the rejuvenators resulted in
moresignicant morphological changes than the aging effect;
however,the rejuvenated blends did not always reproduce the
microstruc-tures as exhibited on the surfaces of the virgin
binders. In addition,the aromatic extract had a stronger impact on
the microstructuralproperties of the binders than the WV oil. The
development ofasphalt binders microstructures depends on the
binders chemis-try and the complicated molecular interactions among
the differ-ent chemical components and any additives [22]. This
helpsexplain the different surface morphological properties in the
bin-der-rejuvenator combinations examined here.
3.2.2. Microscopic mechanical properties of ABD-based bindersThe
surface mechanical properties of ABD based binders were
measured using PFT QNM mode to further explore the
physico-chemical mechanism responsible for the rejuvenation
process.Results of the mechanical properties for virgin, aged, and
aromaticextract rejuvenated ABD samples are shown in Fig. 4. The
gray-scales for each of the properties were set to be the same for
easyvisual comparison among the different samples. Maps of the
differ-ent mechanical properties (adhesion, dissipation, and DMT
modu-lus) for the virgin, aged and rejuvenated samples (Fig. 4)
consist ofsimilar patterns as those from AFM intermittent-contact
mode(Fig. 3). These results indicate that morphological
characterizationfor these samples was repeatable and that complete
blendingbetween the binder and the rejuvenator was achieved. The
minordifferences between morphological properties acquired from
thesetwo modes might result from variance during sample
preparation(Section 2.2.5), namely the amount of binder and
annealingprocedure.
AAD binder
Aged + WV oil Virgin Aged Aged + Aromatic extractin ABD
decreased while the quantity increased, and it appearedthat the
large-sized akes in the virgin binder were replaced bysmaller ones
after aging. On the other hand, bee-structures stillappear on the
surface of the aged AAD binders, and morphologicalchanges between
virgin and aged AAD binder were less obviouscompared to the
ABD-based binders.
Introduction of the aromatic extract into the aged ABD
binder(the third row in Fig. 3) produced dispersed elliptical
domains,with bee-structures at the center of the domains. This is a
similareffect as Pauli et al. observed for the wax-doped binder
samples[22]. The aromatic extract is a petroleum-derived
rejuvenatingagent, and its chemistry supposes to be close to that
of the maltenefractions in asphalt binders. Therefore, the
mechanism responsiblefor the formation of the bee-structures might
occur during therejuvenation process. The aromatic extract
rejuvenated AAD sam-ple exhibited ner-sized bee-structures and the
amplitude ofthe undulated bee-structures was smaller than that of
the virginand aged ones as indicated by the less noticeable
contrast between74 67 199 950.531 0.480 0.325 0.410
-
135X. Yu et al. / FuelHistograms of maps (Fig. 4) of the
adhesion, dissipation energyand DMT modulus of the virgin, aged and
aromatic extract rejuve-nated ABD samples are shown in Fig. 5.
Bimodal distributionsappeared for each mechanical properties of the
samples due tothe contrast between the dispersed phase (namely the
ake-likeand elliptical structures) and the matrix. However, the
chemicalcomponents that are responsible for the different
microstructuresremain unknown. The ake-like domains observed in
virgin andaged ABD samples showed a smaller adhesion, dissipation
andDMT modulus as compared to its surroundings, the matrix
area.However, the area fractions of the ake-like domains in
virginand aged ABD samples were rather small (the less obvious
peaksin the histogram), and therefore only mechanical properties
ofthe matrix are reported as representative values of the sample
sur-faces. For aromatic extract rejuvenated ABD sample, the
bimodaldistributions of adhesion and dissipation energy from the
dis-persed phase and the matrix were more signicant. The
dispersedelliptical domain possessed a smaller adhesion and
dissipation, yeta slightly larger modulus value than the
surrounding matrix, whichagrees with Fischers work [25]. An
interesting observation wasthat for this rejuvenated sample, the
modulus contrast betweenthe dispersed domains and the matrix ipped
around as comparedto the virgin and aged binders, which needs
further investigation ofthe chemical compositions over these
different domains.
Aging decreased the adhesion by about 10% (16 nN and 14.5 nNfor
virgin and aged ABD, respectively). The introduction of the
Fig. 4. Maps of adhesion, dissipation, and DMT modulus (from
left to right) of virgin (topat room temperature (20 C).(2014)
162171 169aromatic extract brought the adhesion force on the matrix
areaback to the level of the virgin binder while the dispersed
phaseshowed a relatively lower adhesion. Upon aging, the
dissipationenergy also decreased slightly whereas the rejuvenated
sampleshad a similar value of dissipation to that of the virgin
binder.
In Fig. 5(c), DMT modulus of the rejuvenated ABD was smallerthan
both the virgin and aged ones, as could be expected. It is
fore-seen that aging stiffens asphalt binders (as shown by the DSR
mea-surement); however, the PFT QNMmeasurements showed that theaged
ABD had a slightly smaller modulus as compared to the virginone,
and repetitive measurements did not produce the same trendas
captured by DSR tests. This discrepancy might be explained bythe
fact that PFT QNM and DSR are measuring mechanical proper-ties of
asphalt binders at different length scales. It is important tokeep
in mind that the AFM measurements represent the surfaceproperties
of a material that can be different from its bulk proper-ties
tested by DSR. How to relate the surface properties at themicro
scale with macroscale bulk measurement for different mate-rials is
still under careful investigation. In addition, other factorscan
also inuence the accurate measurement of binders modulus.For
instance, the DMT contact mechanics that is embedded into thePFT
QNM mode is proven to be inappropriate for retrieving bind-ers
modulus according to the calculated lambda value as
discussedearlier, similar to the scenario in [26]. For future work,
a blunt tipand JKR model are recommended for more accurate
microscopicmodulus measurements. Another factor affecting
accurate
), aged (middle), and aromatic extract-rejuvenated (bottom) ABD
samples measured
-
135(b)
(c)(a)
170 X. Yu et al. / Fuelquantitative measurement of the modulus
could be the use of theinappropriate tip in terms of the relative
stiffness (the ratio of thetip stiffness over the sample
stiffness). The Tap150A tips fromBruker are recommended for
materials with modulus of 4 MPa0.5 GPa, which might not be stiff
enough for probing binder withhigher modulus (i.e., the aged
binders).
In general, it was shown that, values of the mechanical
proper-ties from microscopic measurements of adhesion and
dissipationwere consistent with the macroscopic DSR results. For
instance,the decrease of the complex viscosity after aging and its
recoveryupon adding the aromatic extract agree with the changes in
thedissipation energy from the PFT QNMmeasurements. This
suggeststhat aging and adding aromatic extract as a rejuvenator had
signif-icant impact on mechanical properties of asphalt binders at
boththe micro and macro scales. In addition, microscopic
mechanicalproperties measurements provide supplementary information
forevaluating the rejuvenation effect than merely
morphologicalcharacterization.
3.3. SARA fractionation
In order to explore how the rejuvenator and different
chemicalfractions in asphalt binders contribute to their overall
mechanicalproperties and the rejuvenation mechanism, SARA
fractionation ofaromatic extract alone, as well as virgin, aged and
rejuvenatedAAD and ABD binders was conducted. As shown in Fig. 6,
SARA frac-tionation suggested that aromatic extract contains 40% of
satu-rates, 47% of aromatics, and the rest are resins, with
trivialamount of asphaltene fraction, which is different from the
manu-factures description. To this end, mass fractions of aromatic
extractobtained by the authors were used for further data analysis
since
Fig. 5. Histogram of adhesion (a), dissipation energy (b), and
DMT modulus (c) ofvirgin, aged, and aromatic extract rejuvenated
ABD.the SARA fractionation method was used consistently for
virgin,aged and rejuvenated samples. For both virgin binders, the
resinfraction accounts for the most in their chemical
compositions.
Changes in the SARA fractions of the ABD-based samples(Fig.
6(a)) were signicant. After aging, the saturate and resin con-tents
did not vary much while the aromatic fraction decreasedfrom 28% to
10%, which led to an overall increase of the asphaltenefraction by
about 12%. The effect of aging on ABD is consistent withQins
observation [27] that aging converts the aromatics into the
(a)
(b)
Fig. 6. SARA fractions of aromatic extract, virgin, aged, and
aromatic extractrejuvenated ABD (a) and AAD (b).
(2014) 162171asphaltenes whereas it does not have strong inuence
on saturatesand resins. Adding the aromatic extract introduced more
saturatesand aromatics which, consequently, lowered the fractions
of resinsand asphaltenes compared to the aged binder. The large
amount ofsaturates contained in the aromatic extract helps explain
the sig-nicant increase in the saturate fraction in the rejuvenated
sam-ples. On the other hand, the rejuvenated sample did
notduplicate the mass fractions of the virgin binder.
Similar to the ABD binder, the aged AAD sample had an increasein
asphaltenes and a decrease in aromatics, as compared to its vir-gin
binder. The addition of the aromatic extract increased saturatesand
aromatics content whereas the asphaltene fraction did notdecrease
much as compared to the aged AAD sample. Also in thiscase the
rejuvenated sample did not duplicate the mass fractionsof the SARA
components in virgin AAD.
The molecular interactions among the different binder
compo-nents and additives are very complex. Some chemical
componentsmight be converted to others when asphalt binders are
subjectedto different treatments such as aging and rejuvenation. In
this case,one cannot expect that mass fraction of each SARA
components inthe rejuvenated binders would be equal to the sum of
each fractionfrom the aged sample and the additives based on the
proportion ofthe two used in the blending.
In light of the above analysis and considering the
macroscopicand microscopic mechanical properties of asphalt binders
ABDand AAD discussed in the previous sections, it is a logical step
toconclude that changes in the chemical fractions among the
virgin,aged, and rejuvenated binders are responsible for the
stiffeningeffect of aging and the rejuvenating effect resulted from
therejuvenators. This agrees with the suggestion of Roberts et al.
[6]
-
X. Yu et al. / Fuel 135 (2014) 162171 171that adding the light
bituminous fractions such as aromatics helpsrestoring the
mechanical properties for RAP binders.
4. Conclusions and future work
With the increasing interest in increasing the RAP content
toproduce more sustainable asphalt pavements, various
rejuvenatingagents have been added into themix design so as to
restore the rhe-ological and mechanical properties of RAP
materials. This is animportant step for service-life extension of
asphalt concrete pave-ments made with higher percentages of RAP. In
order to betterunderstand the rejuvenating mechanisms, two commonly
usedrejuvenating agents, WV oil and aromatic extract, were added
totwo binders from different crude sources. Rheological
propertiesat intermediate and high temperatures were measured using
theDSR test, whereas the BBR test was conducted for low
temperaturemeasurements. Surface microscopic morphological
andmechanicalproperties were tested using different AFM devices and
techniques.Chemical compositions of aromatic extract, virgin, aged,
and reju-venated binders were determined using SARA fractionation.
Fromthe above tests, the following conclusions can be drawn:
(1) The complete blending of the rejuvenators into the
agedbinders restored the rheological properties at intermediateand
high temperature ranges as well as the low temperature,as measured
by both DSR and BBR.
(2) The rejuvenating effect depends on both the rejuvenator
andthe crude source of the binder. On the one hand, for ABD-based
binders, 12 wt.% of WV oil reduced the modulus andincreased the
phase angle by a higher extent than the samedosage of aromatic
extract, which may not necessarily bedesirable. In this respect the
aromatic extract achieved thedesired level of rejuvenation on this
binder as compared tothe WV oil. On the other hand, the
rejuvenating effect of aro-matic extract for the two different
binders was related to thetemperature at which their rheological
properties weremeasured.
(3) The rejuvenation effect can be characterized at the
microscale using AFM. Microscopic morphological and
mechanicalmeasurements represent the surface properties of
asphaltbinders, and they provide complementary information
forstudying the rejuvenating mechanisms.
(4) Aging and rejuvenation led to morphological
changes,depending on the specic binder-rejuvenator combination.The
rejuvenated blends did not always reproduce the
surfacemicrostructures of the virgin binders.
(5) Microscopic measurements on adhesion and dissipation
ofvirgin, aged, and aromatic extract rejuvenated ABD sampleswere
qualitatively consistent with the bulk DSR results.However, further
investigations are required in order tobuild the link between
microscopic and macroscopicmechanical properties of asphalt
binders.
(6) Evaluation of the results from the different
characterizationtechniques points out that changes in the chemical
fractionsof the asphalt binders were responsible for the
stiffeningeffect after aging and the improvement of mechanical
prop-erties with the addition of the rejuvenators.
Future work includes modication of the microscopic
modulusmeasurement using AFM and applying the above
characterizationtechniques for various types and quantity of
rejuvenators in orderto provide insight for optimal design of
RAP-based asphaltpavements.Acknowledgements
The authors would like to thank R. Mallick (WPI), M. Tao
(WPI),T. El-Korchi (WPI), S. Granados-Focil (Clark University), R.
Frank(RAP Technologies), and D. Pelegrino (WPI) for their generous
helpin executing this research, and N.A. Burnham (WPI) for her
detailedsuggestions on AFM data analysis, as well as A. Lojoie
(WPI) for herproof-reading of the manuscript.
References
[1] Copeland A. High reclaimed apshlat pavement use. Info brief.
Washington,D.C.: Federal Highway Administration; 2011.
[2] Copeland A. Reclaimed asphalt pavement in asphalt mixtures:
state of thepractise. Washington, DC: Federal Highway
Administration; 2011.
[3] Zaumanis M, Mallick R. Review of very high content reclaimed
asphalt use inplant produced pavements: state-of-the-art. Int J
Pavem Eng, vol. 2014, doi:10.1080/10298436.2014.893331.
[4] Nahar S, Qiu J, Schlangen E, Shirazi M, van de Ven M,
Schitter G, et al. Turningback time: rheological and
microstructural assessment of rejuvenatedbitumen. In: 93rd Annual
meeting of the transportation research board,Washington, D.C.;
2014.
[5] Karlsson R, Isacsson U. Material-related aspects of asphalt
recycling-state-of-the-art. J Mater Civil Eng 2006;18(1):8192.
[6] Roberts F, Kendhal P, Brown ER, Lee DY, Kennedy T. Hot mix
asphalt materials,mixture design and construction, Third Edition
ed., Lanham, MD: NationalAsphalt Pavement Association Research and
Education Foundation; 2009.
[7] Jones DR. SHRP materials reference library: asphalt cements:
a concise datecompilation. Washington, D.C.: Strategic Highway
Reseach Program; 1993.
[8] Zaumannis M, Mallick RB, Frank R. Evaluation of different
recycling agents forrestoring aged asphalt binder and performance
of 100% rectcled asphalt. MaterStruct 2014.
[9] Zaumanis M, Mallick RB, Frank R. Evaluation of rejuvenators
effectivenesswith conventional mix testing for 100% reclaimed
asphalt pavement mixtures.Transp Res Rec 2013;2370.
[10] Zaumanis M, Mallick R, Frank R. Evaluation of different
recycling agents forrestoring aged asphalt binder and performance
of 100% recycled asphalt.Mater Struct, vol. 2014, doi:
10.1617/s11527-014-0332-5.
[11] National Renderers Association Inc, A buyers guide to
rendered products,, Alexandria, VA, 2008.
[12] Masson J, Leblond V, Margeson J. Bitumen morphologies by
phase-detectionatomic force microscopy. J Microsc
2006;221(1):1729.
[13] Yu X, Burnham NA, Mallick RB, Tao M. A systematic AFM-based
method tomeasure adhesion differences between micron-sized domains
in asphaltbinders. Fuel 2013;113:4437.
[14] Hertz H. On the contact of elastic solids. J Reine Angew
Math1881;92(110):15671.
[15] Derjaguin B. Effect of contact deformations on the adhesion
of particles.Progress in surface science. Prog Surf Sci
1994;45(1):13143.
[16] Johnson KL, Kendall K, Roberts A. Surface energy and the
contact of elasticsolids. Proc R Soc A 1971;324(1558):30113.
[17] Maugis D. Adhesion of spheres: the JKR-DMT transition using
a Dugdalemodel. J Colloid Interf Sci 1992;150(1):24369.
[18] Corbett L. Composition of asphalt based on generic
fractionation using solventdeasphalteneing elution-adsorption
chromatography and densiometriccharacterization. Anal Chem
1969;41:5769.
[19] Lesueur D. The colloidal structure of bitumen: consequences
on the rheologyand on the mechanisms of bitumen modication. Adv
Colloid Interf Sci2009;145(12):4282.
[20] Mallick RB, El-Korchi T. Pavement engineering. Boca Raton:
Taylor & FrancisGroup; 2009.
[21] Lu X, Isacsson U. Effect of ageing on bitumen chemistry and
rheology. ConstrBuild Mater 2002;16:1522.
[22] Pauli A, Grimes RW, Beemer AG, Turner TF, Branthaver JF.
Morphology ofasphalts, asphalt fractions and model wax-doped
asphalts studied by atomicforce microscopy. Int J Pavement Eng
2011;12(4):291309.
[23] Soenen H et al. Laboratory investigation of bitumen based
on round robin DSCand AFM tests. Mater Struct
2014;47(7):120520.
[24] Allen RG, Little DN, Bhasin A. Structural characterization
of micromechanicalproperties in asphalt using atomic force
microscopy. J Mater Civil Eng2012;24:131727.
[25] Fischer H, Stadler H, Erina N. Quantitative
temperature-depending mapping ofmechanical properties of bitumen at
the nanoscale using the AFM operatedwith PeakForce TappingTM mode.
J Microsc 2013;250(3):2107.
[26] Dokukin ME, Sokolov I. Quantitative mapping of the elastic
modulus of softmaterials with HarmoniX and PeakForce QNM AFM modes.
Langmuir2012;28(46):1606071.
[27] Qin Q et al. Field aging effect on chemistry and rheology
of asphalt binders andrheological predictions for eld aging. Fuel
2014;121:8694.
Rheological, microscopic, and chemical characterization of the
rejuvenating effect on asphalt binders1 Introduction2 Materials and
methods2.1 Materials2.1.1 Virgin asphalt binders2.1.2
Rejuvenators
2.2 Methods2.2.1 Binder aging2.2.2 Addition of rejuventors2.2.3
Dynamic shear rheometer (DSR) measurements2.2.4 Bending beam
rheometer (BBR) measurements2.2.5 Atomic force microscopy (AFM)
measurements2.2.6 SARA fractionation2.2.7 Test plan
3 Results and discussion3.1 Rheology3.1.1 Properties at
intermediate and high temperatures3.1.2 Properties at low
temperature
3.2 Microscopic morphology and mechanical properties using
AFM3.2.1 Microscopic morphology3.2.2 Microscopic mechanical
properties of ABD-based binders
3.3 SARA fractionation
4 Conclusions and future workAcknowledgementsReferences
