Sie sind auf Seite 1von 26

This article was downloaded by: [FU Berlin]

On: 14 May 2015, At: 15:14


Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer
House, 37-41 Mortimer Street, London W1T 3JH, UK

Road Materials and Pavement Design


Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/trmp20

Asphalt Oxidation Mechanisms and the Role of


Oxidation Products on Age Hardening Revisited
a

J. Claine Petersen & Ronald Glaser


a

Asphalt Materials Consultant , Loveland, CO, 80537, USA E-mail:

Western Research Institute , 365 North 9th Street, Laramie, Wyoming, 82072, USA Email:
Published online: 20 Dec 2011.

To cite this article: J. Claine Petersen & Ronald Glaser (2011) Asphalt Oxidation Mechanisms and the Role of Oxidation
Products on Age Hardening Revisited, Road Materials and Pavement Design, 12:4, 795-819
To link to this article: http://dx.doi.org/10.1080/14680629.2011.9713895

PLEASE SCROLL DOWN FOR ARTICLE


Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of
the Content. Any opinions and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied
upon and should be independently verified with primary sources of information. Taylor and Francis shall
not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other
liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions

Asphalt Oxidation Mechanisms


and the Role of Oxidation Products
on Age Hardening Revisited

Downloaded by [FU Berlin] at 15:14 14 May 2015

J. Claine Petersen* Ronald Glaser**


* Asphalt Materials Consultant
Loveland, CO 80537, USA
clainepetersen@aol.com
** Western Research Institute
365 North 9th Street, Laramie
Wyoming, 82072, USA
rglaser@uwyo.edu

ABSTRACT.

The correlation between asphalt viscosity increase on oxidative aging and the
carbonyl compounds formed (almost exclusively ketones) has been well established; however,
the effect of sulfoxide formation on physical properties during age hardening has received
little attention. Evidence is presented in this paper that shows that the alcohols, which are
formed concurrently with the sulfoxides from the same hydroperoxide precursors as the
ketones, have a similar effect on asphalt viscosity increase as does the ketones. These
alcohols are the main contributors to oxidative age hardening of high sulfur asphalts.
Analysis of the kinetic data for ketone and sulfoxide formation during asphalt oxidation also
provides additional evidence for the validity of the dual asphalt oxidation mechanism
previously reported.
Asphalt Alcohols, Oxidation Kinetics, Ketones, Sulfoxides, Asphalt Oxidation,
Oxidation Mechanisms.

KEYWORDS:

DOI:10.3166/RMPD.12.795-819 2011 Lavoisier, Paris

Road Materials and Pavement Design. Volume 12 No. 4/2011, pages 795 to 819

796

Road Materials and Pavement Design. Volume 12 No. 4/2011

Downloaded by [FU Berlin] at 15:14 14 May 2015

1. Introduction
The chemical and physicochemical mechanisms of asphalt oxidative age
hardening, and changes in the performance-related properties of asphalt resulting from
oxidation, have been the subject of numerous research investigations and much
speculation for many decades. Interpretive reviews of the fundamental work in this
area related to paving asphalts have been published (Petersen, 1984; Robertson, 2001;
Petersen, 2009). The present paper deals primarily with the mechanisms involved in
the formation of ketone and sulfoxide chemical functional groups on oxidation and
their role on asphalt physical property changes. Some of the experimental data
reported are from unpublished work conducted during past decades by one of us
(Petersen) and coworkers at the Western Research Institute (WRI) in Laramie,
Wyoming. Kinetic data reported here on the eight SHRP core asphalts were obtained
at WRI during the course of the Strategic Highway Research Program (SHRP)
(Branthaver et al., 1993), and during following research studies in cooperation with
the Federal Highway Administration (FHWA) (Robertson et al., 2001).
The carbonyl chemical functional group, as characterized by infrared (IR)
spectroscopy, has long been used to indicate the level of asphalt oxidation. The
linear relationship between the log viscosity increase and carbonyl formation during
asphalt oxidation has been well established (Lau et al., 1992; Petersen et al., 1993).
The ketone functional group, the major component of the carbonyl IR absorption
region, is formed primarily from the oxidation of benzyl carbons in side chains
attached to highly condensed aromatic ring systems (Dorrence et al., 1974; Petersen,
1998). These components exist largely in the so-called asphalt polar aromatics
fraction (Corbett-type separation). Ketone formation has been identified as a major
factor leading to asphaltene formation on oxidation, and asphaltenes have been
shown to be primarily responsible for viscosity increase on aging (Petersen, 1984;
Lin et al., 1995). The ketones formed on oxidation have been shown to concentrate
in the asphaltene fraction (Petersen, 2009). It has been proposed that the ketone
functional group per se is not inherently responsible for the viscosity increase, but that
its formation sufficiently changes the polarity, and thus the solubility, of the associated
condensed aromatic ring components to cause them to agglomerate and become part of
the asphaltene fraction, thus increasing viscosity (Petersen et al., 1996).
The other major oxidation product readily identified by infrared spectroscopy is
the sulfoxide functional group; however, the effects of sulfoxides on physical
properties have received little attention in the past, even though sulfoxides are often
produced in significantly greater amounts than ketones, particularly in asphalts of
high sulfur content (Petersen and Harnsberger, 1998; Petersen, 1998). The lack of
interest in pursuing the role of sulfoxides on physical properties has been, in part,
the result of the thermal decomposition of the sulfoxides formed during the high
temperatures of oxidative aging tests often used in the past, making sulfoxide
concentration an unreliable surrogate of oxidation kinetics. Also, of data from
previous studies cited in Petersen and Harnsberger (1998), and Petersen (2009)

Asphalt Oxidation Mechanisms and Age Hardening

797

showed that the asphalt sulfoxides per se had only a small effect on viscosity
increase. However, current modeling studies at WRI (Glaser et al., 2009) have
provided strong evidence that sulfoxide formation is an important factor in
understanding and predicting asphalt oxidative age hardening. These modeling
studies, together with re-evaluation of past WRI research, have led to an expanded
and more fundamental understanding of the role of sulfoxide formation on physical
properties during oxidative age hardening, which is the subject of this paper.
2. Dual asphalt oxidation mechanism

Downloaded by [FU Berlin] at 15:14 14 May 2015

2.1. Typical kinetics of the asphalt dual oxidation mechanism


The kinetics of the major oxidation products produced and the corresponding
viscosity increase for SHRP study asphalt AAB-1 during Pressure Aging Vessel
(PAV) oxidation at 80C and 2.03 x 106 Pa (20 atmosphere) pressure are shown in
Figure 1. All asphalts exhibit somewhat similar kinetics, with an initial rapid
reaction followed by a slower, constant rate reaction. Ketones and sulfoxides are
formed at different rates for different asphalts, and the ratio of ketones to sulfoxides
is highly source dependent (Branthaver et al., 1993). The apparent relationship
between the sum of ketones-plus-sulfoxides and log viscosity increase on oxidative
aging is shown in the figure to emphasize the important, though long unrecognized,
role of sulfoxide formation on viscosity increase, which is discussed in this paper.

Figure 1. Kinetics of ketone plus sulfoxide formation and viscosity increase during
the dual oxidation mechanism for SHRP asphalt AAB-1, PAV aged at 80C
The hydrocarbon chemical reactions dominating the fast and slow reaction
periods have been shown to be fundamentally different (Petersen, 1998). This
previously unrecognized difference in the two different mechanisms is no doubt the

798

Road Materials and Pavement Design. Volume 12 No. 4/2011

primary reason that earlier investigators were unable to unravel the chemistry of
asphalt oxidation. Previous investigators considered only one fundamental
mechanism (Knotnerus, 1971; van Gooswilligen et al., 1985). A discussion of this
subject can be found elsewhere (Petersen, 1998, 2009).
2.2. Chemistry of the dual mechanism of asphalt oxidation

Downloaded by [FU Berlin] at 15:14 14 May 2015

A condensed version of the chemistry of the dual asphalt oxidation mechanism


germane to the discussions in this paper is shown in Figure 2. Details supporting the
mechanisms shown have been published (Petersen and Harnsberger, 1998; Petersen,
1998). More details of the chemistry supporting the mechanism are found in the
Petersen and Harnsberger paper.

Figure 2. Major reaction routes for the dual asphalt oxidation mechanism
Based on the proposed mechanism for the long term asphalt oxidation reaction
shown in Figure 2, ketones and sulfoxides are produced from a common transitory
hydroperoxide oxidation product formed on benzylic carbons (structure VII).

Asphalt Oxidation Mechanisms and Age Hardening

799

Because ketone formation correlates with log viscosity increase, and the chemical
mechanism predicts that alcohol chemical functionality is formed from the benzylic
carbon functionality that forms ketones, it is logical to conclude that the formation
of these alcohols, which are more polar than ketones, should have a significant
effect on viscosity increase. Providing evidence for this hypothesis is a major
objective of the present paper.

Downloaded by [FU Berlin] at 15:14 14 May 2015

2.2.1. Fast (spurt) reaction


In the initial fast (spurt) reaction, oxygen reacts directly with highly reactive
hydrocarbons, argued to be perhydroaromatics (I, Figure 2; Petersen, 1998), to yield
a hydroperoxide (II, Figure 2). The perhydroaromatics oxidation reaction requires no
catalyst, has no induction period, and is estimated, based on past work (Knotnerus,
1971), to be at least 50 times more energetic than the slower hydrocarbon free
radical reaction (Petersen, 1998; Petersen and Harnsberger, 1998). It is this fast
reaction that produces the first order nature of the kinetic plot for asphalt oxidation.
The hydroperoxide produced is quite thermally unstable and can decompose to form
free radicals; however, it readily reacts with alkyl or aryl-alkyl sulfides in asphalt to
produce sulfoxides. The fast reaction also aromatizes the non-planer, strained ring of
the perhydroaromatic precursor, thus increasing aromaticity and introducing
increased planarity and pi-electron cloud density in the ring system. These changes
in structure increase the ability of the molecule to enter into pi-electron associations
of aromatic clusters, which could increase viscosity. Because of the thermally
instability of the hydroperoxide, its decomposition to form free radicals can initiate
the slower, free radical hydrocarbon reaction during the fast reaction time period. At
moderate ambient temperatures, this thermal decomposition also accelerates the
slower, long term reaction during the time period of the fast reaction.
2.2.2. Slow (long term, constant rate) reaction
The less energetic, slower hydrocarbon reaction is responsible for most of the
oxidative age hardening in asphalts during long term pavement service, which can
lead to the deterioration of asphalts performance-related properties. Common
pavement failure modes resulting from oxidation are cracking and raveling. The
long term reaction requires a free radical initiator. Although no induction period is
evident in typical asphalt kinetic plots (often cited as missing evidence for support
of a free radial chain reaction), evidence has been shown for an induction period that
is masked by the rapid fast reaction (Petersen, 1998). The significance of the long
term reaction as it affects asphalt oxidative age hardening is that it involves
oxidation at the benzyl carbon (V, Figure 2) of molecules primarily in the polar
aromatics fraction. These reactive molecules are pre-asphaltene components. The
hydrogen attached to a benzyl carbon is easily removed to form a free radical, which
reacts readily with atmospheric oxygen to form a hydroperoxide (VII, Figure 2).
This hydroperoxide can then react or decompose by several routes; the
preponderance of each route depends on a variety of asphalt source-dependent and

800

Road Materials and Pavement Design. Volume 12 No. 4/2011

Downloaded by [FU Berlin] at 15:14 14 May 2015

environmental variables such as sulfur content, physicochemical microstructuring


(molecular mobility), temperature and oxygen availability.
The principle reaction routes of present interest are (1) the decomposition of
hydroperoxides to form ketones, and (2) the reaction of hydroperoxides with sulfides
to form sulfoxides, as shown in Figure 2. Of pragmatic significance is the fact that
during the formation of sulfoxides, a benzyl alcohol functional group (IX, Figure 2)
is formed. As previously suggested, ketone formation leads to the formation of
components that produce asphaltenes, and thus increase asphalt viscosity. Because
alcohol chemical groups are more polar dipoles than ketones, and have strong
hydrogen bonding functionality, they should be as effective in increasing viscosity
as ketones. Also, because more alcohols are produced than ketones in high sulfur
asphalts than ketones, it follows that alcohols should be a major contributor to
asphalt age hardening in these asphalts. Providing evidence for this hypothesis is the
subject of the following section. Such evidence should also add validity of the
chemical mechanisms previously proposed for asphalt oxidation.
Before considering the evidence, however, it is important to note that during
long term oxidation, the alcohols formed concurrently with sulfoxides are different
than the alcohols formed during ketone formation. The alcohols produced together
with the ketones are primary alcohols formed from the relatively small hydrocarbon
side chains on aromatic rings. Because of their small size and monofunctionality,
they should have little effect on viscosity. This is because when they interact with
another polar component of the asphalt, they effectively cap off the functional
entity; thus, reducing its ability to participate in large viscosity-producing molecular
agglomerates. On the other hand, the alcohols produced during the sulfoxide
reaction are multifunctional because they are part of large molecular agglomerates
having many polar sites. Model studies by one of us (Petersen) showed that
monofunctional polar molecules usually reduced viscosity; and dipolar molecules,
which couple asphalt molecular agglomerates, generally increased viscosity.
3. Evidence for alcohol formation during long term asphalt oxidation
The evidence presented here for the formation of alcohols during long term
oxidation was obtained primarily from data, mostly unpublished, that were collected
during past research at WRI.
3.1. Derivatization of the proposed alcohols using silylation reactions
The asphalts from the California Zaca-Wigmore field experiment, designed to
investigate the role of asphalt source on field performance (Hveem et al., 1959),
were also studied by WRI (Davis and Petersen, 1967). The poorest performing
asphalt, Asphalt E, from this pavement study was derivatized by silylation reactions.
These results have now been found useful in identifying the formation of alcohols in

Asphalt Oxidation Mechanisms and Age Hardening

801

Downloaded by [FU Berlin] at 15:14 14 May 2015

asphalts. Results and are shown in Figure 3, which are tracings from the original
infrared spectra.

a)

b)

c)
Figure 3. Infrared spectral evidence for alcohols in oxidized Zaca-Wigmore asphalt E

802

Road Materials and Pavement Design. Volume 12 No. 4/2011

Hexamethyldisilazane (HMDS) was the silylating reagent used. It reacts with the
hydrogen atoms of OH groups in alcohols, forming the corresponding silyl ethers.
Reactions were usually run up to two to four hours under reflux in 10 ml of carbon
tetrachloride containing 0.067 g asphalt and 0.10 ml HMDS. Following reaction,
Infrared (IR) spectra were obtained on the reaction mixture using 1.0 mm sealed
cells and solvent compensation. The reaction with alcohols is illustrated as follows.
[(CH3)2Si]2NH + 2 ROH 2 [(CH3)2Si]2OR + NH3

Downloaded by [FU Berlin] at 15:14 14 May 2015

The spectra shown in Figure 3 for oxidized Asphalt E components were


separated by pentane precipitation following the asphalts thin film oxidation in the
laboratory at 130C using the IGLC column oxidation procedure (Davis and
Petersen, 1966). In this procedure, asphalt coated on a perfluorocarbon support is
oxidatively aged by passing air through the GC column.
The spectra in Figure 3a were obtained on the asphaltenes separated from neat,
unoxidized Asphalt E. The IR absorption peaks at about 3 600 cm-1 and 3 480 cm-1
result from the free phenolic and pyrrolic chemical functional groups initially present
in the asphalt (Petersen, 1986). The weak, broad band centering near 3 300 cm-1 is a
common peak shared by the hydrogen bonding of both phenolic and pyrrolic
hydrogen. This common peak has been identified and repeatedly observed during
functional group analyses (Petersen, 1986). The IR bands for free OH stretching
vibrations are found in the 3 600 cm-1 region, and the broad band for OH hydrogen
bonding is found in the 3 570-3 200 cm-1 region (Coates, 2000; Jones, 1959).
Alcohol functional groups in asphalts, however, should not be expected to show
significant hydrogen bonding with each other (alcohol dimers) for two reasons.
First, the alcohols exist at relatively low concentrations in oxidized asphalts, and
will be further diluted by the low concentration of sample in the spectral solvent.
Second, there are stronger basic dipoles in asphalts than alcohol oxygen, such as
basic nitrogen, with which the alcohol can hydrogen bond. Such hydrogen bonding
should be expected to shift a portion of the hydrogen bonding band to even lower
frequencies than for neat alcohol self-bonding. A band has been reported in
the 3 600-3 500 cm-1 region for the OH stretch-pi (aromatic) complex (Jones, 1959;
Figueroa et al., 1966; Osawa et al., 1967). Both phenols and benzyl alcohols would
be expected to absorb in this region. Finally, silylation of the OH group of alcohols
and phenols should eliminate free or bonded bands in the IR region just discussed,
and thus serve as confirmation of the reactive OH. Also, the need to distinguish
the IR bands of phenols (which occur naturally in asphalt and are not oxidation
products) from those of the benzyl alcohols formed on oxidation is obvious.
With this background, consider the spectra in Figure 3a. Because silylation had
little effect on the OH infrared region for asphaltenes from unoxidized Asphalt E,
this result provides evidence for insignificant amounts of alcohols in this fraction.
The band at about 3 615 cm-1 provides evidence of phenolics. It is noteworthy that
most phenolics apparently did not silylate, even though phenolics are reactive

Asphalt Oxidation Mechanisms and Age Hardening

803

Downloaded by [FU Berlin] at 15:14 14 May 2015

toward silylation. This observation indicates that the asphaltene component that
contains the phenolics in Asphalt E still exists as highly associated agglomerates in the
carbon tetrachloride solution, making the phenolics less accessible to the HMDS.
HMDS is a rather nonaggressive silylation reagent when used without a co-reactant
such as trimethylchlorosilane to consume the ammonia formed. The small, broad band
at about 3 540 cm-1 most likely arises from phenol-aromatic pi-electron bonding.
Comparison of the data in Figure 3b for oxidized asphaltenes before silylation with
data for unoxidized asphaltenes (Figure 3a) show an almost complete loss of the free
phenolic and pyrrolic functional groups at 3 615 cm-1 and 3 480 cm-1, respectively,
with a large increase in the broad hydrogen bonding region from 3 500-3 100 cm-1
from oxidation. First, note the broad increase in absorbance at 3 500-3 600 cm-1. This
increase is tentatively assigned primarily to the hydrogen bonding of the OH of
alcohols produced on oxidation with the pi-electrons of the condensed aromatic ring
systems in the asphalt. A significant amount of the increase in absorption in the 3 5003 100 cm-1 region is also tentatively assigned to hydrogen bonding of the alcoholic OH
formed on oxidation, together with hydrogen bonding of the native phenolic OH and
pyrrolic NH groups. Rational for these assignments is as follows.
Silylation caused a large reduction in the hydrogen bonding IR bands in those
regions just discussed, with the reappearance of part of the free pyrrolic NH peak
(Figure 3b). First, consider the broad absorption region between 3 500-3 600 cm-1 in
which OH-bonding of phenols and alcohols with aromatic pi-electrons occurs. This
absorption band was completely eliminated upon silylation. Because the formation of
significantly large amounts of new aromatics are not expected on oxidation, increased
phenol-aromatic pi-bonding cannot account for the large absorbance increase in this
region (compare Figures 3a and 3b). Thus, alcohols formed during oxidation are
tentatively identified with the absorbance increase in the 3 500-3 600 cm-1 region. The
loss of the free phenolic peak at 3 615 cm-1 upon oxidation can account for in part of
the large increase in the 3 500-3 100 cm-1 hydrogen bonding region. However, the
concentration of native phenolics (phenolics are not produced on dark oxidation) is
too small to account for the large decrease in the broad hydrogen bonding band on
silylation. Hydrogen bonding of pyrrolic NH with oxidation products could
contribute significantly to the broad increase in the 3 500-3 100 cm-1 region;
however, pyrrolic NH does not silylate under the conditions imposed, and its partial
liberation on silylation is evidence that it was hydrogen bonded to an oxidation
group that silylated almost certainly alcohols. These studies provide strong
evidence for alcohol formation on oxidation, and support for the oxidation
mechanisms shown in Figure 2.
Although free carboxylic acids can absorb in the 3 500-3 600 cm-1 region, they
are virtually excluded from consideration because while conducting extensive
functional group analyses at WRI over the years significantly large amounts of these
acids have not been found to be produced during dark oxidation; the small amounts
formed remain strongly hydrogen bonded, and have lower, broad IR frequencies

804

Road Materials and Pavement Design. Volume 12 No. 4/2011

under conditions used to obtain the IR spectra shown in Figure 3. Further, the band
is much too broad to be considered a free OH peak.

Downloaded by [FU Berlin] at 15:14 14 May 2015

Finally, consider the spectra for the polar aromatics fraction recovered from
oxidized Asphalt E in Figure 3c. Based on criteria applied to the spectra in Figure 3b,
only trace amounts of alcohols are present in this fraction. This result indicates that
virtually all of the alcohols formed on oxidation of Asphalt E were recovered in the
asphaltene fraction during the asphaltene separation. This supports the earlier
conclusion, made when discussing the tertiary alcohols produced during sulfoxide
formation, that the alcohols, together with the ketones, are formed on pre-asphaltene
components of the maltenes and largely become asphaltene components. These data
also further support the proposed mechanism that ketones and sulfoxides (with their
accompanying alcohols) are produced from the same precursor, hydroperoxides,
which are formed on benzyl carbons of condensed aromatic ring structures.
3.2. Additional infrared evidence for alcohol formation
The infrared absorption bands for C-OH functionality formed on oxidation will
be considered next and compared with the OH hydrogen bonding absorption bands
for further confirmation of the formation of alcohols. The data for this comparison
were obtained from a kinetic study of the 130C thin-film oxidation of a Boscan
asphalt using a previously reported column oxidation procedure (Davis and
Petersen, 1966).

Figure 4. Changes in the infrared absorption bands of a Boscan asphalt during


thin-film oxidation at 130C

Asphalt Oxidation Mechanisms and Age Hardening

805

Downloaded by [FU Berlin] at 15:14 14 May 2015

Infrared absorption bands for the O-H bending of tertiary alcohols have been
reported to be in the 1410-1310 cm-1 range, and the C-O stretch region for tertiary
alcohols is reported as a broad band that centers at about 1 150 cm-1 (Coates, 2000).
It is apparent from examination of the data in Figure 4 that oxidation of the Boscan
asphalt produced strong absorption in these two regions. The strong sulfoxide band
at 1 030 cm-1 is superimposed on the broad absorption in the 1 150 cm-1 region;
however, the broad absorption band in the alcohol region is nevertheless readily
apparent.
The absorption bands of the Boscan asphalt attributed to benzyl alcohols
produced on oxidation, and evident in both the 3 200 cm-1 hydrogen bonding band
(centering at about 3 170 cm-1) and the C-OH region (centering at about 1 250 cm-1),
are compared in Figure 5. Data in the figure show a relatively good correlation
between the two absorbance frequencies considering that the absorbance values
were estimated visually from spectra recorded in the 1970s.

Figure 5. Relationship between infrared bands for the proposed alcohols formed on
oxidation in a Boscan asphalt
Based on the chemical reaction mechanism shown in Figure 2, tertiary alcohols
are formed concurrently with the sulfoxides during the long term oxidation reaction.
Thus, one should expect a correlation between the amount of sulfoxides formed and
the net increase in the IR absorbance attributed to tertiary alcohols in the 3 170 cm-1
region. That such a relationship exists is demonstrated from data in the next two
figures. The kinetics of ketone and sulfoxide formation together with the
corresponding increase in the absorption in the 3 170 cm-1 region is plotted in
Figure 6. These data are then cross-plotted in Figure 7 to show the relationship
between sulfoxide formation and increase in the 3 170 cm-1 region for alcohols. Note
that Figure 7 is not a kinetic plot.

Downloaded by [FU Berlin] at 15:14 14 May 2015

806

Road Materials and Pavement Design. Volume 12 No. 4/2011

Figure 6. Kinetics of ketone and sulfoxide formation and the IR absorbance


for alcohols at 3 170 cm-1

Figure 7. Correlation between sulfoxides formed and increases in IR absorbance


attributed to alcohols as predicted by the chemical mechanisms
First, consider data in Figure 6. Most of the chemistry of the fast reaction is
finished after the first hour of oxidation. At the relatively high oxidation temperature
of 130C, the chemistry of the long term reaction proceeds simultaneously and
aggressively with that of the fast reaction. During the next seven hours of oxidation,

Asphalt Oxidation Mechanisms and Age Hardening

807

Downloaded by [FU Berlin] at 15:14 14 May 2015

sulfoxides are being formed from oxidation of sulfides, and also being lost through
thermal decomposition, the latter reaction being significant at 130C (Petersen,
2009). After eight hours of oxidation, the sulfoxide concentration reaches a pseudoequilibrium concentration, with the rate of formation equal to the rate of thermal
decomposition; however, it should be remembered that alcohols are still being
formed during this period even though the effective sulfoxide concentration remains
constant. After significant amounts of ketones have formed during high temperature
aging, their rate of formation decreases because of a corresponding increase in the
rate of formation of dicarboxylic anhydrides and carboxylic acids (nor reported
here), which are formed from the same hydroperoxide precursors that form ketones.
Finally, to show how the relationship between the proposed alcohols and the
formation of sulfoxides can be rationalized with the chemical mechanisms shown in
Figure 2, consider the data in Figure 7 for sulfoxide formation plotted versus the
change in the 3 170 cm-1 IR absorbance attributed to alcohols. During the first hour
of oxidation, the rate of formation of sulfoxides is relatively high relative to the
increase in 3 170 cm-1 alcohol absorbance. This is because, in addition to the
sulfoxides that are formed from the chemistry of the long term reaction that forms
alcohols, significant amounts of sulfoxides are also being formed from the reaction
of sulfides with the perhydroaromatic hydroperoxides of the fast reaction that does
not produce alcohols (see Figure 2). Following completion of the fast reaction, and
up to eight hours of oxidation, there is a linear relationship between sulfoxide
formation and increase in the IR band attributed to alcohol formation, consistent
with the chemistry of the long term reaction shown in Figure 2. Then, after eight
hours of oxidation, sulfoxide formation rate matches decomposition rate; the
sulfoxides reach a pseudo-equilibrium state and the sulfoxide concentration remains
constant even though alcohols continue to be formed as explained above and
evidenced in Figure 7. Finally, at about 24 hours of oxidation, the supply of reactive
sulfides available to form sulfoxides (and thus alcohols) becomes exhausted, at
which point both the formation of sulfoxides and an increase in the 3 170 cm-1
alcohol band cease. This agreement between the chemical mechanisms, infrared
spectra and oxidation products formed provides significant confirmation for the
formation of alcohols and for the chemical mechanisms previously proposed.
4. Chemical functional group-viscosity relationships
As previously mentioned, the role of the ketones formed during oxidation on
physical property changes has received considerable study; however, the role of
sulfoxides has not. Having argued in this paper that the alcohols formed during
sulfoxide formation should have significant effects on viscosity increase, evidence
for this proposal will next be presented. The role of asphalt sulfur content as it
affects the chemistry-physical property relationships during oxidative age hardening
will also be considered in detail.

808

Road Materials and Pavement Design. Volume 12 No. 4/2011

4.1. Low temperature asphalt oxidation

Downloaded by [FU Berlin] at 15:14 14 May 2015

The fast (spurt) oxidation reaction in asphalts proceeds very rapidly at nearambient temperatures as evidenced by data in Figure 8 for a Cosden-Wood River
asphalt aged in air at 45C as a thin film on Ottawa sand. These data were obtained
at WRI in 1986 on uncatalyzed control asphalts that were part of a study on metal
catalyzed asphalt oxidation. Their present value is that they reveal information on
the kinetics of the fast reaction that is difficult to obtain at higher aging temperatures
because of the extremely rapid reaction rate.

Figure 8. Reaction kinetics and functional group-viscosity relationships of asphalt


oxidation at the relatively low temperature of 45C
It is evident that very few ketones were formed at 45C during the time period of
the fast reaction. At this relatively low temperature, the rate of thermal
decomposition of the perhydroaromatic hydroperoxides to form free radicals
(Figure 2) is greatly reduced. The near absence of initial ketone formation is
evidence for an induction period for the long term, free radical hydrocarbon reaction
of the dual mechanism, since no free radicals are formed from the chemistry of the
perhydroaromatic-sulfide chemistry of the fast reaction that forms sulfoxides. As
shown in the kinetic plot on the left in Figure 8, there was a rapid, simultaneous
increase in both sulfoxides and viscosity during the initial stages of oxidation, with a
rapid leveling off when the fast reaction neared completion. Because few ketones
were produced during this period, these results suggest that the rapid viscosity
increase was related to sulfoxide formation. This suggestion is also supported by the
data in the plot on the right in Figure 8.
There is, however, another reaction besides sulfoxide formation that could
explain part of the initial rapid viscosity increase. Because no alcohols are formed
from the fast reaction, the aromatization of perhydroaromatics in addition of the
formation of sulfoxides could contribute to the rapid viscosity increase. Because
perhydroaromatic molecules are strained, non-planer ring systems, they are unlikely
to be significant participants in the strong, pi-bond interactions of aromatic ring
systems that are believed to contribute to molecular agglomeration and thus

Asphalt Oxidation Mechanisms and Age Hardening

809

Downloaded by [FU Berlin] at 15:14 14 May 2015

significantly increase the apparent molecular weight. However, as these molecules


are aromatized during oxidation, their aromatic ring system becomes planar, thus
facilitating aromatic pi-bonding; such aromatization, considered next, could
contribute significantly to the viscosity increase during the fast reaction.
To estimate the relative contributions of the two suggested viscosity-increasing
effects just discussed, first consider the possible viscosity increase from the
sulfoxides produced during the fast reaction. During the SHRP studies, WRI
researchers were concerned about the potential effects of sulfoxide decomposition
on rheological properties during the time that asphalt samples were being annealed
prior to making rheological measurements. Because previous work by one of us
(Petersen) showed that no new oxygen-containing IR bands of potentially viscosityincreasing functionality appeared during sulfoxide thermal decomposition in a
Boscan asphalt, a thermal decomposition experiment was conducted on previously
oxidized SHRP asphalt AAK-1, also a Boscan asphalt, to evaluate the effects of
sulfoxide decomposition on viscosity. A value of 1.3 log Pas/molL-1 was found for
the sulfoxide viscosity effect in asphalt AAK-1 (Petersen and Harnsberger, 1998).
Asphalt AAK-1 is a highly polar asphalt with considerable microstructure; thus, it is
highly probable that the effects of sulfoxides on viscosity increase are greater in this
asphalt than for most asphalts used in industry.
Using this AAK-1 data from the SHRP work, approximate calculations based on
data in Figure 8 for the Cosden-Wood River asphalt showed that if all the viscosity
increase during the fast reaction period, when few ketones were produced, were due
to sulfoxide formation, the effect of sulfoxides on viscosity would be about 6 log
Pas/molL-1, considerably greater than the 1.3 log Pas/molL-1 found for the sulfoxide
effect in asphalt AAK-1. Therefore, these calculations strongly suggest that
aromatization of polar aromatics during the fast reaction must contribute significantly
more to viscosity increase than does the sulfoxides produced. This is not surprising,
since it has been shown that most all sulfoxides are produced on components that do
not form viscosity-building asphaltenes on oxidation (Petersen, 2009).
4.2. Effects of sulfur content on relationships between oxidation products
and viscosity increase during asphalt oxidative aging
4.2.1. Reactive sulfur content in SHRP core study asphalts
The following oxidation kinetic data were obtained at WRI on the eight SHRP
core study asphalts. Because the benzyl alcohols produced in asphalts during the
long term oxidation reaction are produced concurrently with the oxidation of
reactive sulfides, the maximum oxidizable sulfur in the SHRP asphalts, obtained by
three independent methods, are shown in Table 1 to aid in the interpretation the
viscosity-oxidation product relationships. The oxidizable sulfur types are primarily
alkyl or alkly-aryl sulfides; most non-oxidizable sulfur is found in cyclic thiophenic
compounds.

810

Road Materials and Pavement Design. Volume 12 No. 4/2011

Table 1. Maximum oxidizable sulfur in SHRP core asphalts

Downloaded by [FU Berlin] at 15:14 14 May 2015

Percent oxidizable sulfur as determined by


Asphalt

Total sulfur

Quantitative
infrared1

t-Butyl
hydroproxide2

X- Ray
spectroscopy2

AAM-1

1.2

28, 21

17

13

AAG-1

1.3

39

33

34

AAC-1

1.9

31

25

25

AAF-1

3.4

31

28

22

AAB-1

4.7

27

31

21

AAA-1

5.5

29

40

29

AAK-1

6.4

37

36

26

AAD-1

6.9

32

46

38

1. Data from M. Harnsberger, Western Research Institute; 2. Data reported by T. Mill in


Branthaver et al. (1993).

The t-butyl hydroperoxide determination shown in Table 1 was made by reacting


asphalt samples with the hydroperoxide in a reaction similar to the reaction that
occurs during the air oxidation of the native sulfides in asphalts. The X-ray method
used to obtain the data in the last column of Table 1 was sulfur K-edge X-ray
spectroscopy (XANES). The quantitative infrared data shown were obtained at WRI
using functional group analysis (Petersen, 1986).
The data for Functional Group Analysis (FGA) reported in Table 1 are of
particular significance. These data were obtained by analyzing the sulfoxide content
of the SHRP asphalts following 1 000 hours of aging in pure oxygen at 60C
and 20 atmospheres (2.03 MPa) pressure. The relatively good agreement of these
data with the reactive sulfide data obtained by the other two independent methods
show that during the 1 000-hour oxidation period, all of the reactive sulfides were
converted to sulfoxides. This good agreement further serves as an independent check
on the relative accuracy of the FGA method for determining the sulfoxide content of
oxidized asphalts.

Asphalt Oxidation Mechanisms and Age Hardening

811

4.2.2. Comparison of property changes of high and low sulfur asphalts during PAV aging
The kinetic aging data reported in this section were obtained from data available
on asphalts selected from the eight SHRP study asphalts. Summarized chemical and
physical property data for these asphalts have been reported (Jones, 1995). The data
on oxidation kinetics obtained using both the Thin Film Accelerated Aging Test
(TFAAT) and the Pressure Aging Vessel (PAV) have also been reported
(Branthaver et al., 1993; Robertson et al., 2001). Results of subsequent kinetic
studies that followed the SHRP-funded work and used in the present paper are also
available (Petersen et al., 1986; Petersen and Harnsberger, 1998; Petersen, 1998,
2009).

Downloaded by [FU Berlin] at 15:14 14 May 2015

The kinetics of ketone and sulfoxide formation and viscosity increase during the
oxidative aging of a low and a high sulfur asphalt selected from the eight SHRP
study asphalts are compared in Figure 9.

Figure 9. Effects of low and high sulfur content on the ratios of ketones and
sulfoxides formed and their relationship to the viscosity of the oxidized asphalt

812

Road Materials and Pavement Design. Volume 12 No. 4/2011

Downloaded by [FU Berlin] at 15:14 14 May 2015

Asphalts AAM-1 and AAA-1 have sulfur contents of 1.2% and 5.5%,
respectively. The PAV aging data in Figure 9 were obtained at 20 atmospheres
(2.03 MPa) pressure and at 80C, which temperature is near the high end of the
temperature range for asphalt pavements.
Before proceeding with interpretation of the data in Figure 9, a detailed
explanation of what happens during the time periods of the fast and slow reactions is
warranted. At the temperature of 80C during the fast reaction time period, both the
fast (spurt) and slow (long term) chemical reactions are occurring simultaneously. The
long term free-radical chemical reaction occurring during this period appears to be
significantly accelerated by free radicals produced from thermal decomposition of part
of the perhydroaromatic hydroperoxides produced by the fast chemical reaction;
therefore, all possible oxidation products from both the fast and slow chemical
reactions of the dual mechanism are being formed. As previously proposed, the
oxidation products associated with significant viscosity increase are the ketones, the
alcohols produced during sulfoxide formation from the long term reaction, and the
aromatization of perhydroaromatics. The sulfoxides produced from both the fast and
slow chemical reactions, and the small, primary alcohols produced simultaneously
with ketone formation, as previously discussed, have only a small effect on viscosity
increase. It is again emphasized that alcohols formed simultaneously with sulfoxides
by the chemistry of the long term reaction can also be formed during the time period of
the fast reaction. Thus, the chemistry-physical property relationships during the fast
reaction time period are complex, and highly dependent on asphalt sulfur content and
other factors such as temperature and molecular associations (microstructure). As a
result, the linear relationships observed between log viscosity and increase in the
formation of ketones and sulfoxides observed during the long term reaction time
period should not necessarily be expected to extrapolate through the time period of the
fast reaction; this is evidenced by the data for the asphalts shown in Figure 9.
However, when the combination of factors governing viscosity-functional group
relationships fortuitously coincides with the viscosity-functional group effects of the
long term aging period, a linear correlation through both reaction time periods is
found. Another factor that could contribute to the apparent wide differences in these
relationships among asphalts is the inability to accurately determine the changes in
carbonyl concentration at different levels of oxidation. Such an effect, however, should
lead to a rather constant bias between asphalts, and is further judged to be too small to
account for the wide differences in the slope variations observed among asphalts from
different sources.
Returning to Figure 9, compare the kinetic data for ketone, sulfoxide and net log
viscosity increase on aging in the two plots at the left. Both ketone and sulfoxide
formation are accelerated during the fast reaction time period (about 50 hours). During
the subsequent long term reaction period, sulfoxide formation for asphalt AAM-1
ceases; however, for asphalt AAA-1, sulfoxide concentration continues to increase.
For both asphalts, both log viscosity and ketone concentration continue to increase at a
linear rate, albeit at a slower rate during the period of the long term oxidation reaction.

Asphalt Oxidation Mechanisms and Age Hardening

813

Downloaded by [FU Berlin] at 15:14 14 May 2015

To aid in the visual interpretation of these results, ketones, sulfoxides, and


ketones plus sulfoxides concentrations are plotted versus log viscosity increase in
the charts at the right in Figure 9 (note that these are not kinetic plots). The
inflection point in the ketones-plus-sulfoxides plot corresponds with the end of the
fast reaction. Note particularly that during the fast reaction period, the reactions
resulting in the formation of sulfoxides and ketones produce a much smaller increase
in viscosity than is apparent during the long-term oxidation period that follows. The
primary reason for this, as has been discussed, is that although sulfoxides are
produced from the chemistry of both the fast and long term reactions during the time
period of the fast reaction, those sulfoxides formed by the chemistry of the fast
reaction produce only water instead of viscosity-building alcohols.
Next, consider the data for asphalt AAM-1 in the upper right in Figure 9. Based
on data in Table 1, between 13% and 28% (0.05 and 0.10 mol/L) of the sulfur in the
low sulfur asphalt AAM-1 is oxidizable. Thus, all of the oxidizable sulfur in this
asphalt is consumed during the fast reaction period by reaction with the
hydroperoxides of the perhydroaromatics. As a result, only ketones are formed
during the following long term reaction. Because sulfoxides are not being formed
during the long term reaction period in AAM-1, ketone formation drives the
viscosity increase.
Now, consider the data for the high sulfur asphalt AAA-1 shown in the lower
right corner of Figure 9. Between 29% and 40% (0.50 and 0.69 mol/L) of the sulfur
in this asphalt is oxidizable, although only about 0.25 mol/L is consumed during the
fast reaction period; thus, an excess of reactive sulfide remains and is available for
reaction during the long term reaction period. Based on the reaction scheme in
Figure 2, the hydroperoxides formed on benzyl carbons during the long term
reaction period can either decompose to form ketones or react with sulfides to form
sulfoxides. Data in Figure 9 show that ketone formation in AAA-1 is much slower
than sulfoxide formation during the long term reaction period. It follows that most of
the hydroperoxides formed on benzylic carbons are being scavenged by the excess
sulfides rather than decomposing to form ketones. If, as proposed, the alcohols
formed simultaneously with sulfoxides during this reaction produce asphaltenes that
increase viscosity, as do ketones, then the viscosity increase in AAA-1 during the
long term reaction period, under the aging conditions imposed, results largely from
the alcohols produced by the sulfide-sulfoxide reaction. Thus, the sulfoxide-alcohol
reaction dominates in increasing viscosity for high sulfur asphalts, as opposed to
domination by the ketone-forming reaction for low sulfur asphalts.
4.2.3. Examples of viscosity-functional group relationships for high sulfur asphalts
with significant microstructure
Two examples in which the log viscosity-functional group relationships have the
same slope during both the fast and slow reaction periods, as discussed in the
previous section, are shown in Figure 10. Both asphalts AAK-1 and AAD-1 are high
sulfur asphalts with considerable microstructure.

Downloaded by [FU Berlin] at 15:14 14 May 2015

814

Road Materials and Pavement Design. Volume 12 No. 4/2011

Figure 10. Examples where log viscosity-functional group relationships are the
same during both fast and slow reaction time periods
Note that, as apposed to high sulfur asphalt AAA-1 (Figure 9), that considerably
less sulfoxides are produced during the fast reaction period for these asphalts, thus
creating conditions to produce similar log viscosity-functional group relationships
during both fast and slow reaction periods. The effects of microstructure provide a
plausible explanation for this result. Reactive sulfides should be concentrated in the
non-structured phase of the asphalt and the perhydroaromatics in the structured
phase; thus, the logistics of hydroperoxides of perhydroaromatics coming in contact
with reactive sulfides to facilitate the fast reaction, which produces sulfoxides with a
small effect on viscosity, is hindered. This would slow the rate of the fast chemical
reaction and allow more time for perhydroaromatic hydroperoxides to decompose to
form free radicals rather than react with sulfides; the free radicals could thus
accelerate the chemistry of the slow reaction that is occurring simultaneously. The
net result would be the production of a relatively greater amount of viscositybuilding ketones and benzyl alcohols during this period.
4.2.4. Functional group-viscosity relationships during PAV aging for asphalts with
intermediate sulfur content
Figure 12 shows the effects of functional group formation on the viscosity
increase during PAV aging for SHRP asphalt AAF-1, which has a total sulfur
content of 3.4%. This sulfur content is near mid-range for the sulfur content of
asphalts used for paving.
Based on data in Table 1, the calculated reactive sulfide sulfur in AAF-1 is
about 0.30 mol/L. Thus, it is apparent from the kinetic plot in Figure 12, that after
about 100 hours of PAV aging at 80C, virtually all of the reactive sulfides have
been oxidized to sulfoxides. It has been found in a limited number of low
temperature oxidations, in which sulfoxides were formed during the fast reaction

Asphalt Oxidation Mechanisms and Age Hardening

815

Downloaded by [FU Berlin] at 15:14 14 May 2015

with very limited ketone production, that the typical amounts of available
perhydroaromatics in these asphalts were surprisingly similar at about 0.15 mol/L.
Assuming this value for AAF-1, one might expect two inflection points in the plot
for sulfoxide production one when all the perhydroaromatics have been consumed,
and a second when all of the reactive sulfur has been consumed. Data in the plot at
the right in Figure 11 do indeed suggest two inflection points. Based on the earlier
discussion of the chemistry of the fast reaction, significant amounts of sulfoxides are
produced that do not produce a significant viscosity increase. When the
perhydroaromatics are depleted, all sulfoxides that are produced are simultaneously
accompanied by viscosity-increasing benzyl alcohols, thus accounting for the first
inflection point in the sulfoxide plot. Finally, when all of the reactive sulfides are
depleted, no more sulfoxides are produced, accounting for the second inflection point.

Figure 11. PAV aging of SHRP Asphalt AAF-1 (3.4% sulfur content) at 80C
Finally, consider the plot line for ketones at the right in Figure 11, which is
completely linear. A plausible explanation for this, based on the dual mechanisms
shown in Figure 2, is that whenever ketones are produced, any accompanying
sulfoxide production is associated with the formation of benzyl alcohols, which are
proposed to have a similar effect on viscosity as ketones. Thus, when sulfoxide
production ceases, the ketone production increases correspondingly (keep in mind
that his is not a kinetic plot), because both ketones and sulfoxides are produced from
the same hydroperoxide precursor during the long term reaction. Thus, the slope of
the ketone-log viscosity increase plot remains constant.
4.2.5. Effects of pressure on viscosity-functional group relationships on aging
In Figure 12, the effects of atmospheric pressure aging of asphalt AAB-1 at 85C
are compared with PAV aging (20 atm.) at 80C. The two aging temperatures are
believed to be sufficiently similar so that the effects of pressure (oxygen
concentration) can be observed.

Downloaded by [FU Berlin] at 15:14 14 May 2015

816

Road Materials and Pavement Design. Volume 12 No. 4/2011

Figure 12. Effects of pressure on aging kinetics and viscosity-functional group


relationships of SHRP asphalt AAB-1
Asphalt AAB-1 is an asphalt with a significant amount of microstructure in this
aging temperature range. A comparison of the kinetic plots on the left in Figure 12
shows that rate and level of oxidation during the fast reaction time period is greater
for PAV aging than for atmospheric pressure aging. This is no doubt because
pressure increases the oxygen concentration within the microstructure of the asphalt
where the perhydroaromatics are concentrated. Because of the large difference in
reactivity of the precursors for the fast and slow reactions, the fast reaction
dominates if a greater supply of oxygen is present. Note also the difference in the
viscosity-functional group relationships in the plots on the right in Figure 12.
Following the fast reaction, and at a given net increase in log viscosity, it appears
that the amount of ketones-plus-sulfoxides (and thus alcohols) is greater for PAV
aging than for atmospheric aging, again suggesting that relatively more oxidation
occurs deeper in the microstructure during PAV aging where it has less effect
viscosity increase.

Asphalt Oxidation Mechanisms and Age Hardening

817

Finally, it is significant to note that for all asphalts with relatively high sulfur
contents, sulfoxides are produced in much greater amounts than ketones. Thus, if the
alcohols produced during the sulfoxide reaction have a similar effect on viscosity
increase as to ketones, alcohol formation in these asphalts is the major factor
responsible for viscosity increase.

Downloaded by [FU Berlin] at 15:14 14 May 2015

5. Conclusions
Based on the analyses of data presented in this paper, the following conclusions
were reached:
Derivatization of oxidized asphalts together with analyses using infrared
spectroscopy provided convincing evidence that alcohols are formed in asphalts as
predicted by the oxidation mechanism;
Analysis of effects of asphalt oxidation products on viscosity strongly support
the previously proposed dual oxidation mechanism, which predicts that the ketones
and sulfoxides formed during the long term aging reaction are formed from the same
hydroperoxide precursor; thus, when reactive asphalt sulfides scavenge the
hydroperoxides before they decompose to form ketones, viscosity-building alcohols
are produced;
Analysis of the kinetic data for ketone and sulfoxide formation showed that the
alcohols produced have a similar effect on viscosity increase as do the ketones; thus,
these alcohols are the major factor responsible for viscosity increase during the
oxidation and age hardening of high sulfur asphalts;
Finally, the kinetic studies, together with the discovery of the effects on
viscosity of alcohols formed on oxidation, provide convincing confirmation of the
dual oxidation mechanism previously reported.
Acknowledgment
The authors are indebted to both the Western Research Institute and Federal
Highway Administration for their continued support; it is their research efforts and
support that made possible the obtaining of much of the data used in this paper.
Many thanks are also given to Michael Harnsberger and Janet Wolf, through whose
dedicated and tireless efforts the kinetic data were obtained.
Disclaimer
This document is disseminated under the sponsorship of the Department of
Transportation in the interest of information exchange. The United States
Government assumes no liability for its contents or use thereof. The contents of this
report reflect the views of Western Research Institute, which is responsible for the
facts and the accuracy of the data presented herein. The contents do not necessarily
reflect the official views or the policy of the U.S. Department of Transportation.

818

Road Materials and Pavement Design. Volume 12 No. 4/2011

6. Bibliography
Branthaver J.F., Petersen J.C., Robertson R.E., Duvall J.J., Kim S.S., Harnsberger P.M.,
Mill T., Ensley E.K., Barbour F.A., Schabron J.F., Binder characterization and
evaluation, Vol. 2: Chemistry, Report: SHRP-A-368, 1993, Strategic Highway Research
Program, Washington, D.C.
Coates J., Interpretation of infrared spectra, a practical approach, Encyclopedia of
Analytical Chemistry, R.A. Meyers, ed., John Wiley and Sons, Ltd., 2000, p. 1081510837.
Colthup N.B., Spectra-structure correlations in the infra-red region, J. Optical Society of
America, Vol. 40, No. 6, 1950, p. 397-400.

Downloaded by [FU Berlin] at 15:14 14 May 2015

Davis T.C., Petersen J.C., An adaptation of inverse gas-liquid chromatography to asphalt


oxidation studies, Anal. Chem., Vol. 38, No. 13, 1966, p. 1938-1940.
Davis T.C., Petersen J.C., An inverse GLC study of asphalts in the Zaca-Wigmore
experimental road test, Proc. Assn. Asphalt Paving Technol., Vol. 36, 1967, p. 1-13.
Dorrence S.M., Barbour F.A., Petersen J.C., Direct evidence for ketones in oxidized
asphalts, Anal. Chem. Vol. 46, 1974, p. 2242-2246.
Figueroa P.H., Roig E., Szmant H.H., The association of sulfoxides with phenols,
Spectrochimica Acta, Vol. 22, 1966, p. 1107-1115.
Glaser R., Petersen J.C., Turner F., Pauli T., Diffusion with Reaction Model for Kinetic
Studies of Oxidative Aging in Asphalt Binders, Petersen Asphalt Research Conference,
Laramie, Wyoming, July 13-15, 2009.
Herrington P.R., Thermal decomposition of asphalt sulfoxides, Fuel, Vol. 74, 1995,
p. 1232-1235.
Hveem F.N., Zube E., Skog J., Progress report on Zaca-Wigmore Experimental Asphalt Test
Project, ASTM Special Technical Publication, No. 277, 1959, p. 3-45.
Jones D.R., SHRP materials reference library: Asphalt cements: A concise data compilation,
Report SHRP-A-645, 1993, Strategic Highway Research Program, Washington, D.C.
Knotnerus J., Oxygen uptake by bitumen solutions as a potential measure of bitumen
durability, American Chemical Society Division of Petroleum Chemistry Preprints,
Vol. 16, No. 1, 1971, p. D-37-D59.
Lau C.K., Lunsford K.M., Glover C.J., Davidson R.R., Bullin J.A., Reaction rates and
hardening susceptibilities as determined from pressure oxygen vessel aging of asphalts,
Transportation Research Record, No. 1342, 1992, p. 50-57.
Lin M.S., Linsford K.M., Glover C.J., Davidson R.R., Bullin J.A., The effects of asphaltenes on
the chemical and physical characteristics of asphalts, Asphaltenes: Fundamentals and
Applications, E.Y. Shen and O.C. Mullins, eds., Plenum Press, N.Y., 1995, p. 155-176.
Osawa E., Kato T., Yoshida Z., Infrared frequency shifts of phenol owing to hydrogen
bonding with substituted aromatics, J. Org. Chem., Vol. 32, 1967, p. 2803-2806.

Asphalt Oxidation Mechanisms and Age Hardening

819

Petersen J.C., Chemical composition of asphalt as related to asphalt durability state of the
art, Transportation Research Record, No. 999, 1984m, p. 13-30.
Petersen J.C., Barbour R.V., Dorrence S.M., Barbour F.A., Helm R.V., Molecular
interactions in asphalts Tentative identification of 2-quinolones in asphalt and their
interaction with carboxylic acids present, Anal. Chem., Vol. 43, 1971, p. 273-277.
Petersen J.C., Quantitative functional group analysis of asphalts using differential infrared
spectroscopy and selective chemical reactions application and theory, Transportation
Research Record, No. 1066, 1986, p. 106-119.

Downloaded by [FU Berlin] at 15:14 14 May 2015

Petersen J.C., Harnsberger P.M., Robertson R.E., Factors affecting the oxidation and agehardening kinetics of asphalt and the relative effects of oxidation products on viscosity,
Preprints, Division of Fuel Chemistry, ACS, Vol. 41, No. 4, 1986, p. 1232-1244.
Petersen J.C., Branthaver J.F., Robertson R.E., Harnsberger P.M., Duvall J.J., Ensley E.K.,
Effects of physicochemical factors on asphalt oxidation kinetics, Transportation
Research Record, No. 1391, 1993, p. 1-10.
Petersen J.C., A dual, sequential mechanism for the oxidation of petroleum asphalts,
Petroleum Science and Technology, Vol. 16, No. 9-10, 1998, p. 1023-1051.
Petersen J.C., Harnsberger P.M., Asphalt aging: a dual oxidation mechanism and its
interrelationships with asphalt composition and oxidative age hardening, Transportation
Research Record, No. 1638, 1998, p. 47-55.
Petersen J.C., A thin film accelerated aging test for evaluating asphalt oxidative aging,
Proc. Assoc. Asphalt Paving Technol., Vol. 58, 1989, p. 220-237.
Petersen J.C., A review of the fundamentals of asphalt oxidation chemical,
physicochemical, physical property and durability relationships, No. E-C140,
Transportation Research Board, Washington, DC, October 2009.
Robertson R.E., Branthaver J.F., Harnsberger P.M., Petersen J.C., Dorrence S.M.,
McKay J.F., Turner T.F., Pauli A.T., Huang S.C., Huh J.D., Taurer J.E., Thomas K.P.,
Netzel D.A., Miknis F.P., Williams T., Duvall J.J., Barbour F.A., Wright C.,
Fundamental properties of asphalts and modified asphalts, Vol. I, Interpretive Report,
Publication No. FHWA-RD-99-212, US Department of Transportation, Federal Highway
Administration, Mc Lean, VA, October 2001.
van Gooswilligen E. H., Berger H, de Bats F. Th., Oxidations of bitumens in various tests,
Proc. European Bitumen Conference, 1985, p. 95-101.

Received: 7 August 2010


Accepted: 11 April 2011

Das könnte Ihnen auch gefallen