CHARACTERIZATION OF ASPHALT BINDERS
BASED ON CHEMICAL AND PHYSICAL PROPERTIES
Jianghua Weit, Jeffrey C. Shull2,John Barak3, and Martin C. Hawley4
ABSTRACT
Asphalt is a complexmixtureof many differenthydrocarbons.The asphalt
constituentsare classified into three categories: oils, resins, and asphaltenes.
The engineeringproperties of asphaltbinders are directlyrelated to the quantity
and quality of asphaitenes presentand the nature of the dispersionmedium, oils
and resins. Upon addition of a polymerto a straight asphalt, these relationships
and thus engineering properties may be altered. It is therefore necessary to
developa protocolfor characterizingand fingerprintingasphaltbinders.
The chemical compositionsand physical properties of straight(AC-5 and
AC-10) and polymer(SBS and SEBS) modified asphaltswere studiedusinghigh
performance gel permeation chromatography (HP-GPC), Fourier transform
infrared spectroscopy (FTIR), dynamic mechanical analysis (DMA), thermal
mechanical analysis (TMA), and differential scanning calorimetry(DSC). The
HP-GPC system was equipped with a photodiode array detector (PDA) and a
liquid chromatography(LC) transformunit which allowed chemical analysis of
each asphalt constituentusingFTIR.
It was concluded that the combination of HP-GPC and FTIR is an
excellent tool for fingerprinting and quality control of polymers and asphalt
binders. The rheologicalpropertiesof an asphalt binder ware good indices for
determiningthe optimumpolymerconcentrationsfor effective modifications. The
DSC results indicatedthat differentasphalt grades have differentlevels of polar
associationsas detected from changes in enthalpy. Polymer modificationalters
these associations. TMA was a fast methodfor determiningthe highestpossible
servicetemperatureof the pavement.
1.
2.
3.
4.
Research Associate,Departmentof Chemical Engineering
Research Assistant,Departmentof Chemical Engineering
Engineer, Michigan Departmentof Transportation
Professor,Departmentof ChemicalEngineering
B100A Research Complex - Engineering, Michigan State University, East
Lansing,Michigan 48824-1326
729
•
• "_I
I •
CHARACTERIZATION OF ASPHALT BINDERS
BASED ON CHEMICAL AND PHYSICAL PROPERTIES
Jianghua Wei
Jeffrey C. Shull
Martin C. Hawley
The Department of Chemical Engineering
The Composites Materials and Structures Center
B100A Research Complex - Engineering
Michigan State University
East Lansing, Michigan 48824-1326
and
John Barak
Michigan Department of Transportation
Secondary Government Complex
8885 Ricks Road
Lansing, Michigan 48909
A Paper Prepared for The
Intemational GPC Symposium '94
June, 1994
730
,f
TABLE OF CONTENTS
Abstract ............................................................................................................... 1
Introduction ......................................................................................................... 1
Background ................................................. :....................................................... 2
AsphaltCharacterization .......................................................................... 2
PolymerModifierCharacterization ........................................................... 3
Polymer Modified Asphalt BinderCharacterization .................................. 4
Materials and Experiments .................................................................................. 4
BasicChemical Characteristicsof Asphalt Binders :........................................... 6
High PerformanceGel PermeationChromatographyAnalysis .................6
FourierTransform Infrared SpectroscopyAnalysis .................................. 7
LiquidChromatographyTransformAnalysis ............................................ 8
Basic Physical Characteristicsof Asphalts ......................................................... 9
RheometricsAnalysis ............................................................................... 9
Thermal MechanicalAnalysis ................................................................. 10
DifferentialScanning Calorimetry ........................................................... 11
Conclusions ...................................................................................................... 11
Acknowledgments ............................................................................................. 12
References ........................................................................................................ 13
Listof Tables ..................................................................................................... 14
List of Figures ................................................................................................... 14
731
W¢i.ShulLandI-lawlev
ABSTRACT
Asphalt is a complexmixtureof many differenthydrocarbons. The asphalt
constituentsare classified into three categories: oils, resins, and asphaltenes.
The engineering propertiesof asphaltbinders are directly related to the quantity
and quality of asphaltenes presentand the nature of the dispersionmedium,oils
and resins. Upon addition of a polymerto a straightasphalt, these relationships
and thus engineering properties may be altered. It is therefore necessary to
develop a protocolfor characterizingand fingerprintingasphaltbinders.
The chemical compositionsand physicalproperties of straight(AC-5 anti
AC-IO) and polymer (SBS and SEBS) modifiedasphalts were studiedusinghigh
performance gel permeation chromatography (HP-GPC), Fourier transform
infrared spectroscopy (FTIR), dynamic mechanical analysis (DMA), thermal
mechanical analysis (TMA), and differential scanning calorimetry (DSC). The
HP-GPC system was equipped with a photodiode array detector (PDA) and a
liquid chromatography(LC) transform unit which allowed chemical analysis of
each asphalt constituentusingFTIR.
It was concluded that the combination of HP-GPC and FTIR is an
excellent tool for fingerprinting and quality control of polymers and asphalt
binders. The theological propertiesof an asphalt binder were good indices for
determiningthe optimumpolymerconcentrationsfor effective modifications. The
DSC results indicated that differentasphalt grades have different levels of polar
associationsas detected from changes in enthalpy. Polymer modification alters
these associations. TMA was a fast method for determiningthe highest possible
service temperatureof the pavement.
INTRODUCTION
The long term pavement performance of asphalt concrete (AC) surfaced
roads is a function of traffic load and volume, material properties, construction
practices, and environmental factors. A typical AC surfaced pavement
deteriorates over time and withincreasingnumberof load repetitions. Pavement
deterioration manifests itself in several common types of distress including
rutting, fatigue cracking, low temperature cracking, reflective cracking, aging,
ravelling, and stripping13. The conventional materials used in the asphalt
concrete mixture may perform satisfactorilyrelative to one distresstype but fail
prematurelyrelative to the others. For example, asphalt concrete mixturesmade
by using hard binders will have low rutting but, high fatigue and temperature
cracking potentials. On the other hand, mixtures made with softasphalt binders
will have low fatigue and temperature cracking but, high rutting potentials.
Hence, modification of the asphalt binder to enhance its performance at both
high and low temperatures and under traffic loading is essential to the success
of constructing superior pavements. Such modifications includethe addition of
polymers to enhance the binder properties at both low and high temperatures.
732
Wei. ShuU.and I-Iawl_-
And indeed, it has been shown that polymer modified asphalts can improve
pavementperformance4_.
Michigan State University is currently conducting a study of polymer
modified asphalt pavements that will address the above considerations. The
study is divided into three sections including, the fundamental physical,
chemical, and thermodynamic properties of asphalt binders, the basic
morphologyand microstructureof polymer-fiber-asphalt-aggregatemixtures,and
the structural and engineering properties of polymer-fiber-asphalt-aggregate
mixturesunder extremelow and hightemperaturesfound in Michigan.
Materials being used in the study include AC-2.5, AC.-5, AC-10, AC-20,
styrene-butadiene-styrene (SBS), styrene-ethylene-butylene-styrene (SEBS),
epoxy, polyethylene (PE), ethylene-vinyl-acetate (EVA), various fibers, and
recycled crumb rubbertires.
This paper focuses on the development of a protocol for fingerprinting,
including characterization and identification,asphalts using high performance
gel permeation chromatography (HP-GPC) and other instrumentation. The
asphalts studied were straight, styrene-butadiene-styrene(SBS) and styreneethylene-butylene-styrene (SEBS) modified AC-5 and AC-10 asphalts. The
molecular information of the binders was studied using HP-GPC, Fourier
transform infrared spectroscopy (FTIR), and differential scanning calorimetry
(DSC).
The physical properties of the binders were studied using DSC,
dynamicmechanical analysis(DMA), and thermal mechanical analysis(TMA).
BACKGROUND
Asphalt Characterization
Asphalt is a complex mixture of many different hydrocarbonsconsisting
primarily of molecules that contain mainly carbon and hydrogen atoms. The
chemical compositionof an asphalt is determined by the source of crude oil,
refinery process, and the grade of the asphalt. A generic asphalt elemental
analysis shows that approximately 84 percent of the sample is carbon, 10
percent hydrogen, 1 percent oxygen,and the remainder consistsof several trace
elements includingnitrogen, sulfur, vanadium, nickel, and ironr. The average
molecularweight of a generic asphaltmoleculeranges from 500 to 50007. Table
I shows basic molecular information of three commonly used asphalts, AC-5,
AC-10, and AC-207.
In general, asphalt constituentsare classified into three categories: oils,
resins, and asphaltenes. Oils are the light compounds in asphalt whichhave the
lowest molecularweights (24 - 800) and have a large numberof side chains and
few rings. Accepted criteria for the oil classification are molecules with
carbon/hydrogenatom number ratios less than 0.6 that are soluble in hexanea.
Resins are intermediate molecular weight compounds (800 - 2000). It is
importantto note that resins can contain sulfur and nitrogen. Resins are polar,
have a carbon/hydrogen ratio between 0.6 and 0.8 and are soluble in light
733
•"
Wei.ShuU.andI-Iawl_petroleum naphtha8. Asphaltenes are the highest molecular weight compounds
(1800 - 8000) with aromatic ring structures, few side chains, and
carbon/hydrogen ratios greater .than 0.8a. Asphaltenes contain the trace
elements mentioned earlier which may react with potential polymers and are
soluble in carbon-tetrachloride.
An average asphalt sample has an
asphaltenelresin/oilweight ratio of approximately231271509and the asphaltene
contentis higher for harder asphalts. HP-GPC is usually used to determinethe
molecular weight, molecular weight distribution, and the fraction of each
constituentin the asphalt.
Using the asphaiteneiresin/oil classifications,a two phase asphalt model
was developed8 and is illustrated in Figure 1. Phase 1 is an assembly phase
consisting of asphaltenes and resins that is dispersed in phase 2, the solvent
phase consistingof the oily constituents. The resins behave as peptizingagents
that stabilizethe asphaitenes in the oily constituents.
The asphaltene core of the assembly phase can vary between 40 and 60
angstroms in diameter. As the asphaltene content of an asphalt sample
increases, the assembly phase also increases. This adds structure to the
asphalt and gives it better high temperature properties, such as increased loss
and storage moduli,resulting in enhanced ruttingresistance. The percentage of
the assembly phase present is the defining factor between different grades of
asphalt. The resins contain aromatic compounds substitutedwith longer alkyis
and a larger number of side chains attached to the ringsthan asphaltenes. The
combination of the saturated and the aromatic characteristics of the resins
stabilizesthe colloidalnature of the asphaltenes in the oil medium. Asphaltenes
are present as discrete or coUoidally dispersed particles in the oily phase.
Colloidaily dispersed asphaltenes are not stable in the oil medium by
themselves, but can be stabilized through polar resins. Asphaltenes can exist
both in a randomly oriented particle aggregate form or in an ordered micelle
form. In miceUeform, the polar groups (water, silica, or metals such as V, Ni,
and Fe) are either oriented toward the center to form an oil-externalmicelle from
hydrogenbonding,charge transfer, or salt formation,or oriented outwardto form
an oil-internal micelle (Hartley micelle). The growth and ultimate size of the
micellesare dependent on temperature, resin content,and the presence of other
chemicals such as polymer modifiers. The engineering properties of asphalts
are directly related to the quantity of asphaltenes, the size of the micelle
structure,and the nature of the dispersionmedium,oils and resins.
Polymer Modifier Characterization
Accordingto the functions and behaviors of various modifiers in asphalts,
modifierscan be categorized into five types: dispersed thermoplastics,network
thermoplastics, reacting polymers, fibers, and crumb rubber (CRM) particles.
Dispersed thermoplastics behave like asphaltenes and normally require
peptizing agents like resins to stabilize the modified systems. Usually, it
requires a considerable amount of material before forming a macrostructural
network. Network thermoplasticsbehave like resins and will form a network of
734
|
1
,
.
.
o •
wei. Shuil, and Hawley
themselves inside asphalts. Reacting polymers bond chemically to the asphalt
(normally to the asphaltenes) and will form asphalt/polymernetworks. Typical
functional groups that may react with polymers include: carboxylic acids,
ketones, phenols, suifoxides, acid anhydrides, pyrroles, and quinones7. CRM
particlesbehave as aggregates if their sizes are large and behave as dispersed
thermoplastics if their sizes are small (below l OOMm). Fibers increase the
available wetting surface area and behave as binder thickenerswhich reduce
asphalt bleeding. In all cases, the goal is to create more structure inside an
asphalt without losingits low temperatureproperties.
Polymer Modified Asphalt Binder Characterization
Asphalt is a viscoelastic material that displays a variety of different
propertiesdepending on the temperature of the sample. Goodrich, proposed a
model of asphalt depicted as a shock absorber and a spring_°. The shock
absorber representsthe viscous properties of asphalt and the spring represents
the elastic properties. At high temperatures, asphalt shows good viscous flow
propertiesand behaves as a shock absorberwith little or no elastic behavior. At
low temperatures, asphalt becomes a "brittle elastic solid" with little or no
viscous properties. In the spring/shockabsorber analogy, the spring becomes
over loaded and snaps when low temperaturecrackingoccurs.
In order to reduce the potential of rutting at high temperatures and
thermal cracking at low temperatures, various polymers/fibers/rubbershave
been added to asphalt. The goal is to increase the temperaturerange of both
the elastic and viscous properties of the asphalt binders and mixtures.
Convenient measuresof these properties are G', storage modulus(elastic), and
G", loss modulus (viscous). The ideal case will be that at high temperatures
both G' and G" increase upon addition of polymer due to a network structure
formation within the asphalt; and at low temperatures,G' and G" decrease upon
addition of polymer due to a decrease in the material'seffective glass transition
temperature. The ideal case leads to the improvementof both viscous and
elastic properties over a wide range of temperatures.
MATERIALS AND EXPERIMENTS
Materials used in this studywere AC-2.5, AC-5, AC-10, AC-20, SBS, and
SEBS. SBS and SEBS were used to modify both AC-5 and AC-10 at 0 to 10
weight percent polymer content. Aged asphalts were also studied. Short term
aging (experienced during processing and the first year of service) was
simulatedby using the Rolling Thin Film Oven Test (AASFTO T 240; ASTM D
2872). Long term aging (experienced during approximately15 years of service
life) was simulatedby usingthe Pressure AgingVessel Test (SHRP B-005).
A Millipore/Waters' high performance gel permeation chromatography
system was used to analyze the molecular weight distributions of straight
asphalts,polymer modifiers, and polymer modified asphaltbinders. The system
735
'
Wei. ShulI. and Hawley.
was equipped with a differential refractometer, a photodiode array detector, a
chromatographymanager (Millenniumsoftware), a dual reciprocatingpump,and
a manual injectionport. A Lab. Connections' liquid chromatographytransform
(LC-Transform) unit was also installed as an additional detector for the system.
The HP-GPC systemwas equipped with four Waters' styragelHR columnsthat
encompassed an effective molecularweight range from 0 to 600,000. HP-GPC
grade tetrahydrofuran at a flow rate of 1.0 mL/min, was used as the mobile
phase for this study.The sample concentrationwas kept constant at 10 grams
per I liter of solventas was the injectionvolumeof 250 mL.
The differentialrefractometer was capable of measuringmolecularweight
distributions of asphalts, polymers, and polymer modified asphalts. The
differential refractometerwas very reproducible. Figure 2 is an overlay of two
AC-10 chromatogramsprepared independentlyfrom each other.
The photodiodearray detector allows for the characterization of useful
chemical functionalgroupspresent in asphalt. PDA was not used extensivelyin
this study, but it will be useful in the future to discriminate between certain
asphalts and polymersbased on their ability to absorb UV light. The axes of
PDA data are time (molecularweight), UV absorbance, and wavelength. Using
the Millennium software package, dissectionof this data is possible that allows
individual spectraand chromatogramsat a fixed molecularweight or wavelength
to be examined.
The LC-Transform is an off-line unit that collects the effluent from the
differential refractometer(separated by molecular weight) on a germanium disc
that can be analyzed using Fourier transform infrared spectrometryat a later
date. The LC-transform/FTIRwill also allow us to characterize useful chemical
functional groupsthat absorb infraredlight.
Fourier transform infrared spectroscopywas used to examine unaged,
aged, and polymer modified asphalts. The samples were prepared by
evaporating a THF/binder solutionon a potassiumbromide salt pellet leaving a
thin film of the materialto be examined. It was determined that the absorption
peaks at 1375, 1450, 1600, and 1700 cm1 correspond to CH3, CH2, aromatic
carbon, and carboxyl groups in the asphalt, respectively. The heights of the
absorption peaks were assumed to be proportional to the amount of the
functional groupspresentin the sample. Figure 3 shows the FTIR spectrumof a
straight AC-5 sample. FTIR was also used to determine the possible
characteristicpeaks corresponding to the functional groups of SBS and SEBS
polymers. Figure 4 showsthe infrared spectrumof SBS in which a unique peak
can be identified at a wavelength of 965 cm"_(There is no 965 cm"1peak in
Figure 3). This peak is assigned to the trans-l,4 contribution of butadiene in
SBS. No characteristic peak was identified in the FTIR spectrum of SEBS.
Clearly, FTIR will be a useful tool for fingerprinting and quality control as well as
determinationof polymercontentfor SBS modified asphalts.
Dynamic mechanical analysis experiments were performed using a
Rheometrics RMS-800 apparatus. All experiments utilized smooth, 25 mm
diameter plates in a parallel arrangementwith a gap width between 1.4 and 2.0
736
T
" W=.ShulLandl_wlw
mm. Sample handling procedures were consistent with those used for the
SHRP (Strategic Highway Research Program) Bohlin instrumentin which the
sample is poured hot directly onto the plates and allowed to cool to room
temperature. Temperature sweeps were then conducted from 25 to 60°C with
measurementstaken at five degree intervalswith an equilibrationperiod of two
minutes. A frequency of 10 radianslsecondwas used (compliance with SHRP
specification)and strain levels were controlled to insure that the testing was
conducted in the linearviscoelasticrange. The strain level was within 0.5 to 9%
for all tests. All tests were replicated between three and six times to insure
accurate and reproducible results.
For example, Figure 5 shows the
reproducibilityof the storage mudulusversus temperaturedata for a triplicate.
Thermal mechanical analysis tests were performed on representative
asphalt binder samples. The samplesware placed in a 5 mm diameter by 5 mm
deep glass boat at room temperature and packed firmly forming a fiat top
surface. The sample was then covered with a 3 mm diameter by 1.5 mm thick
glass plate upon which a one gram load was placed. The sample was then
cooled at approximately 15°C per minute to a temperature of -100°C and
allowed to equilibrate. The temperature was then raised at a rate of 5°C per
minute and its height change was measured until a total melt (correspondingto
a 1 mm depressionof the sample) was achieved. After testing, the boats were
retrieved to insurethat the sample did not flow over the glass cover/plate. All
testswere duplicatedto check the reproducibilityof the results.
Differential scanning calorimetry was used to detect glass transition
temperatures and equilibriumtemperatures of possible molecular associations.
Asphalt sampleswere equilibrated at different temperaturesand then quenched
using liquid nitrogento a temperature of-100oc.
The enthalpy change during
heatingwas then recordedas the samplewas heated at a rate of 10°Clmin.
BASIC CHEMICAL CHARACTERISTICS OF ASPHALT BINDERS
Basic chemicalcharacteristics of asphalt binders were studied using HPGPC, LC-Transform, and FTIR. Results from each study are summarized as
follows.
High Performance Gel Permeation Chromatography Analysis
An HP-GPC chromatogramof a straightAC-5 is shownin Figure 6. Three
distinct molecularweight ranges (divided by lines A and B) are present. These
separations do not necessarily correspond to the molecular weight ranges that
classifythe oil, resin, and asphaltene constituents found in the literature using
solubilitytechniques. However, the reported percentages of the constituentsdo
exist if the molecularweight ranges (divided by lines C and D) reported in the
literature are followed. Table II lists the percentages of the constituents
according to the chromatography separation and separation according to
reported literature values. From HP-GPC chromatograms,the molecularweight
737
,
•
Wei. Shull. and FIawlev
averages and polydispersities of asphalt binders were obtained. Table III lists
the averages for various materials. These data may later be correlatedwith both
physicaland engineeringpropertiesof asphalts.
A comparisonof HP-GPC molecularinformationof straightAC-5 and AC10 is listed in Table Ill. As can be seen, It is difficult at this time to distinguish
between asphalt grades at this time.
A comparison of HP-GPC chromatogramsof aged and unaged AC-5 is
shown in Figure 7. It can be seen that when AC-5 is aged, there is an increase
in the high molecular weight material and a decrease in the low molecular
weight material. During aging there is only approximately 0.2% weight loss
suggestingthat the low molecularweight material is being reacted.
Gel permeation chromatography has also been used to examine the
differences between pure SBS and SEBS. Table III indicates that the SBS
polymer being studied has a weight average molecularweight of approximately
62,000 g/mol and a polydispersityof two. The SEBS polymer being studied on
the other hand has a weight average molecular weight of about 45,000 g/tool
and a polydispersityof 1.35.
Because polymer molecules are larger than asphalt moleculesas shown
in Table III, gel permeation chromatographycan be used as an effective tool for
determining the polymer content of the asphalt/polymerblends being studied.
Figure 8 is an example of this techniquefor AC-5 modified with4 weight percent
SEBS. When the original AC-5 sample is subtracted from the asphalt/polymer
blend, the entire chromatogramcan be divided into three regions:A, B, and C
with corresponding areas of 2.52%, 1.49%, and 95.99 %, respectively. Region
A is the large polymer molecules that are totally separated from the asphalts.
Region B is the small polymer molecules that are eluted with the asphalL This
can be easily seen by comparison of Figure 8 with Figure 7. A total of 4 weight
percent SEBS can be identified from the HP-GPC data. It is importantto note
that it appears the polymer is being degraded, sheared, or reacted during the
mixing process. Similar resultswere obtainedfor AC-51SBS,AC-101SEBS, and
AC-10/SBS asphalt/polymerblends.
A comparison of HP-GPC chromatograms of aged, straight AC-5 and
aged, 4% SBS polymer modified AC-5 is shown in Figure 9. Upon addition of
polymer to asphalt, we would expect to see an increase in the amount of the
high molecularweight asphaltconstituentfor the aged sample due to both aging
and possible polymer degradation. Instead, we see a decrease in this value
suggestingthat the polymerhas reduced the amount of aging that has occurred.
Fourier Transform Infrared Spectroscopy Analysis
FTIR spectroscopywas used to characterize various functional groups in
asphalts and polymermodifiers. Figure 3 is a representationof these groupsfor
an aged AC-5 sample. Ratios of the absorbance intensitiesof these functional
groups can yield useful informationwhen fingerprintingdifferentasphalt grades.
Examinationof absorbance intensityratiosindicatesthat an AC-10 sample has a
greater percentage of aromatic carbon than AC-5 as would be expected
738
Wei. Shrill.andHawley.
suggesting a larger asphaltene content. However, AC-10 appears to have less
CH2 functional groups than AC-5 suggestinga lesser resin content. Peaks at
1375, 1450, 1600, and 1700 cm1 correspond to CHs, CH2, aromaticcarbon,and
carboxyl groups present in the asphalt, respectively. Analysis such as this in
combinationwith gel permeation chromatographywill be useful when selecting
materialsand dealingwith recyclabilityconcerns.
As the FTIR spectrum of SBS has a unique characteristicadsorptionpeak
at a wavelength of 965 cm1, FTIR is an excellent tool for identificationand
determinationof the SBS content in an unknown asphalt binder. To determine
the SBS content using FTIR, the absolute ratio of the 965 cmI adsorption band
to the 1375 cm"1band can be used as a content indicator. This ratio is linearly
proportionalto the SBS content in the asphalt. However, a calibration curve
must first be constructed that is a function of asphalt and polymer type and
source. A calibration curve is constructed by preparing samples that contain
known concentrationsof SBS polymer in the asphalt sample and plottingthis
information against the respective absolute absorbance band ratios. It is
important to note that the calibration curve is valid only for the given
asphalt/polymerblend. If different materials are used, a new calibration curve
must be constructed. The curve must also be corrected for aging discrepancies.
Figure 10 is an example of a calibrationcurve for an unaged AC-51SBSpolymer
blend. The definingequationfor the unaged calibrationcurve is:
Ratio = 0.064*polymer content + 0.037
while that of an aged curve is:
Ratio = 0.051*polymer content + 0.079.
Uquid Chromatography Transform Analysis
As not all polymermodifiers have a characteristic absorption peak that is
unique from that of the asphalt spectrum, FTIR alone may not be sufficientfor
fingerprinting all polymers. However, it will be much easier to identify the
polymersif we can separate them from the asphalt/polymerblends. As polymers
have much higher molecular weights than asphalts (see Table III), polymers
elute from the HP-GPC column first (see Figure 8). By attaching an LCtransform unitto the end of the HP-GPC system, it is possible to collectonly the
polymerportionof a blend on a germanium disc that can be analyzed using an
FTIR system. Figure 11 shows the FTIR spectrum of SBS collected usingthis
technique. The spectrum in Figure 11 shows the same characteristics as those
found in the spectrumpresented in Figure 4 (originalSBS).
By combinationof HP-GPC, LC-Transform, and FTIR, the contentand the
nature of most non-reacting polymers in an asphalt/polymer blend can be
obtained.
739
,
Wci.ShuU.andHawl_
BASIC PHYSICAL CHARACTERISTICS OF ASPHALTS
Basic physical characteristics of asphalts were studied using
Rheometrics, DSC, and TMA. Results from each study are summarized as
follows.
Rheometrics Analysis
The major importance of DMA testing is to show that the presence of
polymer in an asphalt binder does indeed enhance the theologicalpropertiesof
the blend over a wide temperature range. As polymers will enhance asphalt
properties more at high temperatures than at low temperatures, it is properto
focus on the property enhancement at high temperatures. Figure 12 shows the
sensitivityof tan8 with respect to SBS polymer content and temperature for a
modified AC-5 asphalt binder. Clearly, SBS polymers enhance the theological
properties of AC-5, especially at high temperatures. Similar resultswere found
for all blend combinationsstudied in this research includingAC-51SEBS, AC10/SBS, and AC-101SEBS. It is importantto note that an extensive mixingstudy
was first conducted to insure that optimum blend properties were being
measured". The mixingprocedureused called for preheatingthe raw materials
to 135°C for one hour and then mixing with a low shear mixer (approximately
1600 rpm) equipped with a four blade, 5 cm diameter impeller for two additional
hours at 180°C.
It is also important to determine the optimum and/or critical polymer
contentfor best modification.After an extensive study, the optimumwas defined
as the point at which there is no longer a significant increase in the theological
properties of the blend with increasingpolymer content". The critical content,
on the other hand, was defined as polymer concentrationswhere a deviation
from the normal trend in the theological propertieswas detected. The Strategic
Highway Research Program(SHRP) uses llJ" at 60°C as a theologicalproperty
indicatorfor its original binder specification. Figure 13 shows the inverse loss
compliance as a function of SBS polymer content at 60°C for a modified AC-5
asphalt binder. Based on the definitionspresented and Figure 13, the optimum
and critical polymer contentsare 5 and 4 percent for this system. Additional
theological indicators were also investigated that verified the optimum and
critical polymer contents". The same experiments were conducted for AC51SEBS, AC-101SBS, and AC-101SEBS systems. The optimum and critical
polymer contentsare listed in Table IV.
The decrease in properties at the critical polymer content is a real
phenomenon that has been tested multiple times using several experimental
techniques. This phenomenonwas also observed in the TMA analysis as will be
discussed in a later section. A possible explanation of the observed behavioris
that as polymer is added to the asphalt, the thermo-physicalproperties of the
unmodified material are altered due to solubility of the oils in the polymer, or
some other structuralphenomenon. At the same time, the polymer is enhancing
the overall rheologicalpropertiesof the entire blend. It could be envisionedthat
740
W¢i. ShuU, and Hawley
when two rheologicalproperty profilesare added together they yield a resultant
propertyprofile similarto that observed in this study.
The aging effects on the. rheological properties of AC-5/SBS polymer
blends (polymer contents from zero to ten percent) were also studied using
Rheometrics. The samples were aged usingthe Thin Film Oven and Pressure
Aging Vessel Tests. These samples were then tested and the results were
comparedwith those of the unaged samples. No observable trend with respect
to increasing polymer content was observed. However, Figure 14 depicts the
percent increase in G' at 60°C of the aged versus unaged polymer modified
binderfor several SBS polymercontents. Increasingpolymer contentappears to
decrease the magnitudeof increase in the storage modulus. This may suggest
that asphalt aging is retarded throughthe addition of a polymer modifier which
agrees with the conclusiondrawnfrom the HP-GPC analysis.
Thermal Mechanical Analysis
Thermal mechanical analysis tests have shown that the glass transition
temperatures (Tg) of both SBS and SEBS polymers to be in the range from -60
to -90°C. Upon additionof polymerto asphalt, the Tg of asphalt/polymerblends
will be lowered since the Tg's of straight asphalts are much higher than -60"C.
TMA tests were also used to look at the final melt temperatures of the four
asphalt/polymer systems: AC-5/SBS, AC-51SEBS, AC-101SBS, and AC101SEBS. Figure 15 summarizesthe TMA resultsfor the four systemsat various
polymer contents. The TMA resultsindicatethat:
1. All of the TMA data support the conclusionsdrawn for the critical and
optimumpolymer contentsbased on rheological properties (see Table
W).
2. The final melt temperature of AC-10 is greater than that of AC-5 as
expected.
3. At their respective optimumpolymer contents,there is little difference
in the final melt temperaturesof the asphalt/polymerblends.
4. At their respective optimumpolymer contents, there is an increase of
approximately 25°C in the final melt temperature for the
asphalt/polymerblendswith respectto the straightbinder.
The aged AC-51SBS polymer blends were also tested using thermal
mechanical analysis to determinethe final melt temperaturesof the samples as
shown in Figure 16. Examination of Figure 16 indicates that increasing the
polymer content beyond the optimum percentage has an insignificanteffect on
the final melt temperature of the aged AC-51SBSblends. The difference in the
final melt temperatures of the aged versus unaged samples for increasing
polymer content decreases. At approximately nine percent polymer content,
74]
"
Wei. Shulk and Hawley.
there is no difference in the final melt temperatures between the aged and
unaged samples. This polymer content roughlycorresponds with the level of
modificationthought to cause a matrix inversionof the AC-51SBS blend. The
TMA data support the HP-GPC result that addition of polymer retards age
hardening. Also, TMA data show that no age hardening occurs beyond the
matrixinversion(about 9 wt.%).
Differential Scanning Calorimetry
Differentialscanningcalorimetry was usedto measure the glass transition
temperatures of AC-5 and AC-10 asphalts. Figure 17 shows the DSC
thermographof an AC-5 asphalt. The glass transitiontemperaturesof both AC5 and AC-10 are within the range from -10 to -15°C. Upon addition of polymerto
asphalt, the blend's glass transition becomes a broad range between those of
the polymersand asphalts.
Differential scanning calorimetry has also been used to examine any
temperaturedependent structurethat may exist in asphalt. Straight AC-2.5, AC5, AC-10, and AC-20 asphalt sampleswere quenchedfrom room temperatureto
-100°C using liquid nitrogen. The samples were then heated at a rate of
10°C/minute. Figure 18 shows the DSC thermographsfor representative runs.
it has been found that there are three distinct thermal transitions that exist in
straight asphalt at approximately -12, 10, and 33°C. The first transition is
believed to be the glass transition of the asphalt. The second and third
transitionsmay be due to polar associationsbetween asphalt molecules. The
transition temperature corresponds to the breakdown temperature of the
association. This speculation was qualitativelyverified by the fact that if an
asphalt sample was equilibrated at an elevated temperature for a period of one
hour before quenching, the transitionsat 10 and 33°C could be eliminated as
seen in Figure 19. The temperatures in Figure 19 represent the equilibrium
temperature used before the DSC test. After ackee of the AC-5 sample at 350C
for one hour,the transitionat 10°C was eliminatedbut the transition at 33°C still
exists. After ackee of the AC-5 sample at 135 or 180°C for one hour, both
transitionsat 10°C and 33°C were eliminated. Also, this is a reversible process.
If a heated sample is allowed to equilibrate at room temperature, the transitions
will reappear as seen in Figure 20. The reversible phenomenon verifies the
speculation that these transitions are structural changes due to polar
associationsand not chemical transitions. More research is needed in this area
to determine the nature of these transitionsand how they will be affected upon
addition of polymer.
CONCLUSIONS
This study of the characterization of asphalt binders based on chemical
and physicalpropertiesshows the following:
742
Wei, Shull. and Hawley.
•
HP-GPC, LC-transform, FTIR, TMA, DSC, and DMA were used to
measure chemical and physical properties of straight and polymer
modifiedasphaltbinders.
•
HP-GPC was an excellent tool for determiningthe polymer content in an
asphalt binder.
•
A combination of HP-GPC and LC-transform/FTIR can be used for
fingerprintingthe nature of asphalts and polymers as well as quality
control of asphaltsand polymermodified asphaltbinders.
•
DMA results showed that rheoiogical properties were enhanced upon
additionof polymersto asphalt binders, especiallyat high temperatures.
•
DMA can be used to identify the optimum and critical concentrations of
polymers required for producing an effective enhancement of the
rheologicalproperties of an asphalt binder.
•
TMA results show that the melt temperature of a binder increases with
polymercontentand asphalt grade.
•
DSC results suggested that different extents of structure existed for
different asphaltgrades. Addition of polymersaltered these structures.
•
Addition of SBS and SEBS polymersto AC-5 and AC-10 asphalt binders
appears to retard asphalt age hardening.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the financial support provided by the
Michigan Department of Transportation (MDOT) and the technical assistance
given by MDOT, the Composites Materials and StructuresCenter, and the Civil
and Environmentaland Chemical Engineering Departmentsat Michigan State
University.
743
Wei. Shull. and Hawley.
REFERENCES
1. Baladi, G.Y. Fatigue Life and Permanent Deformation Characteristics of
Asphalt Concrete Mixes. In Transportation Research Record 1227, TRB,
National ResearchCouncil,Washington, D.C., 1989.
2. Baladi, G.Y., R.S. Harichandran, and R.W. Lyles. Asphalt Mix Design - An
Innovative Approach. In Transportation Research Record 1171, TRB,
National Research Council,Washington, D.C., 1989.
3. Baladi, G.Y. /ntegrated Material and Structural Design Method for Flexible
Pavements. Final Report RD-88-109 and 110. FHWA, U.S. Department of
Transportation,!987.
4. Newcomb, D.E., M. Stroup-Gardiner, and J.A. Epps. Laboratory and Field
Studies of Polyolefinand Latex Modifiers for Asphalt Mixtures. In Polymer
Modified Aspha# Binders, ASTM ST/= 1108. K.R. Wardlaw and S. Shuler,
Editor. American Society For Testing Materials, Philadelphia,1992, pp. 129150.
5. Rogge, D.F., R.L Terrel, and A.J. George. Polymer Modified Hot Mix
Asphalt - Oregon Experience. In Polymer Modified Asphalt Binders, ASTM
STP 1108. K.R. Wardlaw and S. Shuler, Editor. American Society For
Testing Materials, Philadelphia, 1992, pp. 151-172.
6. Khosla, P.N. Effect of the Use of Modifiers on Performance of Asphaltic
Pavements. In Transportation Research Record 1317, TRB, National
Research Council,Washington, D.C., 1991, pp. 10-22.
7. Material Reference LJ'brary Ashalt Data, in Strategic Highway Research
Program, National Research Council, Washington, D.C. 1992.
8. Highway Materials Engineering - Asphalt Materials and Paving Mixtures. Vol.
FHWA-I-90-008, Washington, D.C. 1990, U.S.D.O.T., Federal Highway
Administration,National Highway Institute.
9. Baladi, G., PersonalCommunication,1992.
10. Goodrich, J. L., What Is The Purpose Of Modifying Asphalt With Polymers?
1990, Chevron Research And TechnologyCompany.
11. Shuil, J. C., Wei, J., Eberz, K., Hawley, M. C., Baladi, G. Y., and L. T. Drzal
The Effect of Polymer Content on the Rheological Properties of AC-51SBS
Blends. Submittedto TRB, National Research Council, Washington, D.C.,
for Annual Meeting, January, 1995.
Wei. Shull. and Hawley
LIST OF TABLES
Table
I
II
III
IV
Caption
Asphaltelemental analysis
Percentage of constituentsin an AC-5 asphalt binder
Molecularweight averagesfor various materials
Optimumand critical polymercontentsfor asphaltmodification
LIST OF FIGURES
Figure
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Caption
Two phase asphalt model
Reproducibility of Refractive Index detector(AC-10)
FTIR spectrumof an aged, unmodifiedAC-5 sample
FTIR spectrumof SBS
Storage modulusas a functionof temperaturefor a triplicateat four
percent SBS polymercontent
HP-GPC raw data for unaged, straightAC-5
Comparisonof molecularweight distributionsbetween aged and
unaged AC-5
HP-GPC raw data for unaged AC-5 modified with 4 percent SEBS
polymer
HP-GPC raw data comparingaged, 4 percent SBS polymer
modifiedand unmodifiedAC-5 binders
FTIR absorbance band ratio calibrationcurve for unaged AC5/SBS polymerblends
FTIR spectrum of SBS from LC-transformseparation
Sensitivityof tan delta with respect to SBS polymercontent
Inverse losscompliance as a function of SBS polymercontent at
60°C
Percentincrease in the storage modulusat 60°C of aged versus
unaged SBS polymer modifiedAC-5 binders
Melt temperaturesof asphalt/polymerblends as a function of
polymercontentdeterminedby TMA
Melt temperaturesof aged and unaged AC-5/SBS polymerblends
as a function of polymercontent determined by TMA
DSC data for determinationof glass transitiontemperatureof AC-5
DSC data depictingintensityof transitionsas a function of asphalt
grade
DSC data depictingtransitionsas a function of temperature history
DSC data depictingtransitionsas a functionof repetitive cycling
745
WCLShull. andHawley'
Table
I Asphalt
AC-5.
85.7
10.6
-0.54
5.4
163
36
w
32.5
7.24
570-890
Cart)on,%
Hydrogen, %
Oxygen, %
Nitrogen,%
Sulfur, %
Vanadium, ppm
Nickel, ppm
Iron, ppm
Aromatic C, %
AromaticH, %
MolecularWeight
(Toluene)
Table II Pecentage
Classification
Asphaltene
Resin
Oil
elemental
of constituents
ChromatographySeparation
MolecularRange
Percent Area
6700 +
7.45
2900 - 6700
11.85
0 - 2900
80.70
Table III Molecular
weight
analysis
AC-10
82.3
10.6
0.8
0.54
4.7
220
56
16
31.9
7.12
810-930
AC-20
84.5
10.4
1.1
0.55
3.4
87
35
100
32.8
8.66
840-1300
in an AC-5 asphalt
binder
LiteratureSeparation
Molecular Range
PercentArea
2300 +
23
950 - 2300
27
0 - 950
50
averages
for various
materials
.................
..N..ame.
..........................
.M....n
............
MP_
............
................
.M....z.
...........
AC-5
AC-10
AC-5 (aged)
SBS
SEBS
AC-514%SBS
AC-514%SEBS
AC-514%SBS (aged)
AC- 10/4% SEBS
AC-10/4% SBS
Table IV Optimum
669
630
792
30002
32730
657
752
803
780
791
and critical
862
922
876
67865
44092
829
871
876
917
907
polymer
2099
1865
3348
61900
44445
4320
4322
5508
4602
6086
content
6725
5660
11121
106729
58327
51605
38754
48220
42458
82759
3.14
2.96
4.23
2.06
1.36
6.57
5.75
6.86
5.90
7.69
for asphalt
modification
optimum content
critical content
(percentby totalweight of the asphalt/polymerblend)
AC-5/SBS
AC-5/SEBS
AC-10/SBS
AC-101SEBS
5
5
2
4
746
4
3- 4
3
3
We_ 5hu_ andHaSty.
phase
2
(assemblmes)
Figure I Two Phase asphalt model
0.70
Figure 2 Reproducibility
of Refractive Index delmctor (AC-10)
747
Wei, Shulk and Hawley
I_
i
i
i
i
i@e
B
D
9
i
•
•
Figure 3 FTIRspectrum of an aged, unmodifiedACE sample
,o
!:.
•19
.le
OI
_
Figm 4 FTIR spectrumof SSS
748
. ,+J
' -
w_ Shun,aadt-L_,y
1000.00
100.00
I
10.00
1.00
0.10
t
30
20
I
40
I
SO
!
60
I
?0
i
|0
Temperature(C]
Figm
5 Storage modulus as a function of temperature for a triplicate at
four pen:m SBSpolymercontent
80.00-
60 ooi
40.00
20.00,
.I
"1_
O. 00-[
I
"
v
_ p
1
24:00 " 2s:oo ' 2s'oo " 30:00 ' 3_':oo ' 34:o0
_s:oo
_s'oo ,io:o0
PU.DU_eS
Figure 6 HP-GPC raw data for unagecl,stmJght AC-8
749
"
Wa. Sh_U,and
0.80-
"
/'""
,
0.40-
\
o.zo-_r_,_ed
0.00,
4.'40 4.'z6 4.00 :_.eo 3.G0 3.40 ",.zo :_.00 ,.eo 2._0 2.40 z.z
loq
MN'
Rgure 7 Comparison of mlecular weight distributions between aged
and __
AC-5
SO0.OG
400.00
i
300.00.
200.00
1-00.00
!1
0.00-
,_
22"00
v
:_4'.00
26.00
28.00
30.00
" 32-'00
34.00
" 36;00
" 38'.00
" 40'.00
'
HI_tce',
Rgm
8 HP-GPC raw data for unaged AC-5 modified with 4 percent
seem
750
600.0
500.00.
200.1
Figm
9 HP_I=C raw data comparing aged, 4 _
modified and unmodified AC-5 binders
SBS polymer
:
0.,
:
0.$
•
o.,
_ 0.3
_
•|01: o.,
0.1
*
l
m
o
•
•
0
l
I
2
,P
3
l
l
I
I
l
i
I
4
S
6
7
8
9
10
Sii$
Weisbt
PercoRt
Figure 10 FTIR absorbance band ratio calibration curve for unaged AC-
sses
bends
I
751
I
•
Wei, ShulLand Hawi_
Figure 11 FTIR spectrum of SBS from LC-transform separation
12
10
-4--0
-4--t
--o,-2
-o..- 3
8
:"
o
="
.
.-o--4
--_--S
s
_10
2
0
20
|
30
I
40
Temperature
Rgm
12 Sens_
!
60
|
SO
I
70
n
80
(eC|
of tan delta with respect to SBS polymer content
752
7
100
1
0
I
I
!
I
t
2
4
6
8
10
Percent
Polymer
Content
Figure 13 Inverse loss complianceas a function of SBSpolymercontent
at 60°C
0
3%
45
Percent
Polymer
%S
Content
Figure 14 Percentage increase in the storage modulus at SO°Cof aged
versus unaged SBS polymer modified AC-5 binders
753
'
w_ sb_, a_ I-I_
120
110
_-. 100
E
a.
oi
u.
E
-;
9O
80
-o--AC-$O w/SBS
/
I -°-AC's
--o-- AC-10 w'w/SEBS}
8BS
1
7O
60
O
I
2
3
4
S ,
ti
7
II
t
10
Percent Polymer Content
i
Figure15 Melttemperatures
of asphalt/polymer
Mendsasa functionof
polymer content determined by TMA
120
110
g
I
|
100
,o
m
t
.S
"-
80
..e.. U aalod
--a--Aged
70
S0
I
2
_
4
6
I
8
i
le
Percen!PolymerConlen|
Figure 16 Melt temperatures of aged and unaged AC-8/SBS polymer
Mends as a function of polymer content determined by TMA
754
0.1
0.0
-la 72-c (Ts)
_
•
-0.4 -
3
_
-0.2
-
-0.3-
_
-50
Rgm
17 DSC dm
AC_
o2_
0
lemlPeratwre
for determination
0.4
1.
755
25
(*C)
of glm
5O
_'wn_W_
_
"75
of
c
J
4
w_, sl_u, _i l-la_.
o.3
*o.J
b_
""-o.3, _
W
t_
0
-
-0.5
- _0
-_5
0
Telper
_._
e f.ur e
_o
;5
("C)
Figure 19 DSC data depicting transitions
histmy
as a function of temperature
|U
8'
S "_
0
-3o
- io
_
Temlllers_r
_
•
5o
70
("C|
Figure 20 DSC data depicting Inmldtionl as a function of repelilive cycling
756