Geotextiles and Geomembranes 42 (2014) 641e647

Contents lists available at ScienceDirect

Geotextiles and Geomembranes

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

Interaction of antioxidants with carbon black in polyethylene using

oxidative induction time methods
Wai-Kuen Wong a, *, Y. Grace Hsuan b, 1
a
GSE Environmental, 19103 Gundle Road, Houston, TX 77073, USA
b
Department Civil, Architectural and Environmental Engineering, Drexel University, 3141 Chestnut Street, Philadelphia, PA 19104, USA

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

Article history: Carbon black (CB) is added to the geosynthetics for ultraviolet protection. However, CB can interact with
Received 25 May 2014 antioxidant (AO) in the geosynthetics, and their interaction has been documented to be synergistic and/
Received in revised form or antagonistic towards the oxidation reaction. In this paper, two oxidative induction time (OIT) methods,
7 July 2014
a standard (Std) OIT test and a high pressure (HP) OIT test, were used to evaluate the interaction of CB
Accepted 23 July 2014
Available online 23 September 2014
with two types of antioxidants during the oven aging. Test samples made from high density polyethylene
(HDPE) blended with different amounts of Irganox®1010 (I-1010) and Irgafos®168 (I-168) together with
0 or 2.5wt.% of carbon black were studied. The sample were incubated in a forced air oven at 85
C to
Keywords:
Carbon black
accelerate the oxidation reactions. After 1000 days, Std-OIT values of samples with 500 ppm and
Antioxidant 1000 ppm I-1010 exhibited only 10% drop from their initial values from 19.42 min and 34.01 min to 17.53
Oxidation and 31.0 min respectively. In comparison, a continuous decrease of Std-OIT value was observed for
Oxidation induction time samples contained both I-1010 and I-168; a 40% drop was measured after 1000 days, corresponding to
Polyethylene 16.4 and 26.8 min reduction. For samples contained 2.5% CB and I-1010, an exponentially decrease of Std-
OIT and HP-OIT with aging time was obtained. However, adding I-168 to those samples did not change
the OIT decreasing trends, indicating that the interaction of CB with I-168 is negligible. The test data
further verify that the effectiveness of the OIT test is strongly correlated to the functional temperature
range of the AO. The Std-OIT test with testing temperature of 200
C can detect both I-1010 and I-168,
while the 150
C testing temperature of the HP-OIT test can only detect the I-1010.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction
Geosynthetics made from HDPE are commonly incorporated
2e3% of carbon black (CB) for ultraviolet (UV) protection together
with approximately 0.1e0.3% of antioxidant (AO) to ensure oxida-
tion resistance of the products during their service time. When
adding CB (channel black or furnace black) to pristine HDPE
without AO, it has been found to delay the onset of oxidation in
polyethylene. The function of suppressing thermal oxidation is
related to the amount of oxygen containing group on CB surface
(Hawkins et al., 1959a, b; Pleshanov et al., 1982). On the other hand,
the degradation mechanism of different types of AOs on protecting
polyolefin products under various environmental conditions, such
as multiple heating cycles of extrusion, radiation, and thermo-
* Corresponding author. Tel.: þ1 281 443 8564; fax: þ1 281 230 8650.
E-mail addresses: conniewkwong@gmail.com (W.-K. Wong), ghsuan@coe.
drexel.edu (Y.G. Hsuan).
1
Tel.: þ1 215 895 2785; fax: þ1 215 895 1363.
oxidation, were well studied (Epacher et al., 1999a, b; Pock et al.,
2004; Moore et al., 1988; Moss and Zweifel, 1989; Drake et al.,
1990; Malle
gol et al., 2001).
Even though AO and CB can protect polymer from oxidation
degradation individually, when combined, both synergistic and
antagonistic effects were found. By increasing CB loading or
reducing the particle size of CB, it can enhance the OIT value of AO
stabilized PE (Phease et al., 2000; Wong and Hsuan, 2012). On the
other hand, Hawkins et al. (1959a) and Kovacs and Wolkober (1976)
measured the oxygen absorption rate and found that CB combined
with certain types of hindered phenol or secondary amines can
greatly reduce the AOs' effectiveness in PE. The acidic nature of CB
due to the high oxygen content, can hinder the performance of AO.
They explained that the quinone active sites on the CB surface are
stronger acceptors than the free radicals, resulting in AO reacting
with the quinone instead of free radicals. Pena et al. (2001a, b) used
flow micro-calorimetry to measure the heat of adsorption/
desorption of primary and secondary AOs when blended with four
types of CB. They found that the AO adsorption and desorption
activities were not only dependent upon the particle size of CB but

http://dx.doi.org/10.1016/j.geotexmem.2014.07.009
0266-1144/© 2014 Elsevier Ltd. All rights reserved.
642 W.-K. Wong, Y.G. Hsuan / Geotextiles and Geomembranes 42 (2014) 641e647
also influenced by oxygen groups content, neighboring surface
functional groups, CB surface structure, and porosity.
In past decade, oxidation resistance of HDPE has been intensely
studied, and oxidative induction time (OIT) test has been used in
many of the studies to assess the remaining AO in samples with and
without carbon black. Subjecting AO stabilized PE samples in aer-
obic and anaerobic environments at elevated temperatures,
Lundback et al. (2006) used the OIT test to monitor the AO loss as
well as to study AO diffusion through the thickness of the sample.
Nagy et al. (2003), investigated the effectiveness of the AOs in
polyethylene and the resistant of AOs to the hydrolysis. Thomas and
Ancelet (1993) studied 14 HDPE geomembranes with different
weight percentages of three types of AOs (Irganox®I-1010, Tinu-
vin®622, Irgaphos®I-168). Samples with and without CB were also
evaluated. They found that the standard (Std)-OIT test cannot
detect Tinuvin®622 (a HALS stabilizer) in samples without CB,
while a significant higher HP-OIT value was measured. Further-
more, the service life of AOs in HDPE containing 2e3% CB have been
investigated by many research teams (Smith et al., 1992; Viebke and
Gedde, 1997, 1998; Hsuan and Koerner, 1998; Rowe and Rimal,
2008; Muller and Jacob, 2003; Wong and Hsuan 2009). All of
these studies provide valuable findings and insights of the perfor-
mance AO in PE products with and without CB. However the
interaction between CB and AO during aging of HDPE products is
not well studied and understand.
This paper focuses on two objectives: i) study the effects of CB
on AO during oven aging through ten AO/CB formulations. ii)
Evaluate the long term oxidation of the two commonly used AO,
Irganox®I-1010 (I-1010) and Irgaphos®I-168 (I-168). To understand
the depletion mechanism of the two AOs, both Std and HP-OIT
methods were utilized to measure the remaining AO in samples
aged in a forced air oven at 85
C. Furthermore, the effectiveness of
the two OIT methods in detecting these two AOs was also evalu-
ated. A long aging period (1200 days) was implemented to cultivate
the complete depletion of one or both AO in the samples.
2. Test material
A total of ten HDPE samples with different amounts of AO and
CB were prepared for this study. The neat HDPE resin used has a
density of 0.953 g/cm3 and melt index of 0.16 g/10 min. The HDPE
resin was blended with Irganox®1010 (I-1010) and Irgafos®168 (I-
168) with four concentration combinations. I-1010 is a type of
hindered phenol serving as processing stabilizer as well as anti-
oxidant during the product's service life. I-168 is a phosphite based
AO and is function as a processing stabilizer. Carbon black (CB),
65 nm particle size, at 2.5 wt% was added to half of the AO blended
samples. Also two referenced samples without AO, one neat HDPE
Table 1
Antioxidant formulations of blended samples.
Sample Antioxidant Carbon black
I-1010 (ppm) I-168 (ppm) (wt.%)
PE0 0 0 0 4. Results
PE0-C 0 0 2.5
PE1 500 0 0
PE1-C 500 0 2.5
PE2 500 1000 0
PE2-C 500 1000 2.5
PE3 1000 0 0
PE3-C 1000 0 2.5
PE4 1000 1000 0 induce free radical reactions and consume part of the AOs even in
PE4-C 1000 1000 2.5
resin and one with 2.5wt% CB, were prepared. The ten formulations
are shown in Table 1.
All samples were first extruded into pellets using the laboratory
scale single screw extruder. Materials without CB were prepared by
first dry blending HDPE resin fluff with an appropriate amount of
AOs and then extruded into pellets. For material with carbon black,
pellets were prepared by mixing the PE fluff, AOs, and CB powder
(grinded from the CB master batch which consisted of 50 wt.% CB
and 50 wt.% linear low density PE) and then extruded into pellets.
The pellets were compression molded into plaques according to
ASTM D4703, A1 at a cooling rate of 15
C/min. The thickness of the
plaques ranged from 1.8 mm to 2.1 mm. Each plaque was then cut
into 75 mm  90 mm test samples for oven incubation.
3. Incubation and OIT test methods
3.1. Oven incubation
All test samples were incubated in a forced air oven (THELCO
Laboratory Oven, Model 70M) at 85
C to accelerate the AO
depletion. Samples were hung on the stainless strain racks inside
the oven with 1 inch spacing. Samples were retrieved at different
time intervals to assess the remaining AO using OIT tests.
3.2. Standard OIT (Std-OIT) test
The Std-OIT test was performed using the TA instrument Q20
Differential Scanning Calorimeter (DSC). The Std-OIT test was car-
ried out according to the procedure described in ASTM D3895. A
thin strip of the specimen was cut across the thickness of the
sample and weighted to 3.5 ± 1 mg.
3.3. High Pressure OIT (HP-OIT) test
The HP-OIT test was performed using TA instrument 2920 Dif-
ferential Scanning Calorimeter equipped with a high pressure cell.
The test was carried out according to the procedure described in
ASTM D5995, using the constant volume procedure. A thin strip of
specimen was cut across the thickness of the sample and weighed
to 3.5 ± 1 mg.
3.4. High Performance Liquid Chromatography (HPLC)
Prior to the HPLC analysis, Soxhlet extraction procedure was
performed to obtain the remaining AO in the sample. Sample were
grinded to 100-mesh particle size (100 mm) and extracted using
cyclohexane, according to the procedure described in ASTM D6082.
The extraction solutions were filtered through 0.45 mm PTFE filters
before the analysis. AGILENT series 1100 HPLC with a UV-detector
was used for HPLC analysis. The antioxidant concentration was
then analyzed using reversed-phase HPLC, with acetonitrile/water
(95/5) as the mobile phase and a Waters Symmetry C18 column
(150 cm length). An injection volume of 10 ml and a flow rate of
0.5 ml min were used. The concentration of antioxidant was
determined from the absorption at 280 nm.
4.1. Initial OIT value
As the compound (HDPE and AOs) is going through the extru-
sion and then compression molded into a plaque, the high pro-
cessing temperature and shear action during the extrusion can
an oxygen deficient environment (Kriston et al., 2009, Moss and
 W.-K. Wong, Y.G. Hsuan / Geotextiles and Geomembranes 42 (2014) 641e647 643

Zweifel, 1989). To quantify the remaining AO concentration after
extrusion and sample plaquing, AOs were extracted from the test
samples and then analyzed by HPLC to determine the remaining
amount of I-1010 and I-168. However, it is necessary to mention
that HPLC can only detect the original structure of I-1010 (none of
the four phenolic groups are reacted during processing). If one or
more of the functional group reacted with the free radical and the
others remain active, HPLC analysis cannot recognized it as I-1010,
even though it may still act as functioning AO. Also, the Std-OIT and
HP-OIT tests were used to determine the OIT value of the unaged
samples. The results of HPLC and OIT tests are shown in Table 2. The
Std-OIT value of PE0 is less than 0.5 min which is similar to the
value of MDPE resin puffs (Hsuan and Koerner, 1998). Since PE0
samples only experienced compression molding, the short period
of heating during the plaquing did not seem to induce oxidation
degradation in the polymer. For PE0-C, the additional of 2.5wt% CB
increased the Std-OIT value to 1.2 min while there was no detect-
able AO from the HPLC analysis. Thus, the increase of the Std-OIT
value was contributed solely from the CB which functioned as a
mild antioxidant in the polymer as observed by Hawkins (1984)
and Pleshanovet et al. (1982). For samples that went through the
extrusion processing, depletion was measured in both AOs. The
consumption of I-1010 and I-168 was less than 30% of their initial
concentrations for all samples except PE2, in which 60% of the I-
1010 and 50% of I-168 were consumed. The reason for higher AO
consumption in this sample was unknown. Comparing the
remaining AO in samples without CB (PE3 and PE4), adding I-168
reduced the consumption of I-1010 for about 10%. In contrast, the
concentration of I-1010 in CB samples decreased about 25% from
their initially added levels. Also a similar I-1010 concentration was
detected in CB samples with the same initial I-1010 amount, i.e.,
PE1-C and PE2-C; PE3-C and PE4-C. Blending I-168 did not prevent
the consumption of I-1010 in CB samples during the extrusion
process.
For the initial OIT value, adding I-168 increased the Std-OIT
value, while its influence on the HP-OIT value was negligible, see
Table 2. I-168 raised the Std-OIT value by about 18 min in the two
pairs of samples with I-1010 and CB samples (PE1-C/PE2-C and PE3-
C/PE4-C). For samples without CB, the increase was about 14 min
and 25 min for PE1/PE2 and PE3/PE4, respectively. It is believed that
the different effects of I-168 on the Std-OIT and HP-OIT values was
caused by the effective temperature range of the AO at which the
maximum reaction rate between antioxidant and free radicals took
place. I-168 is designed to be a processing stabilizer with an
effective temperature ranging from 180 to 300
C, while the
effective temperature of I-1010 ranges from room temperature to
300
C (Fay and King, 1994). The 200
C testing temperature of Std-
OIT falls within the effective temperature range of these two AOs,
and thus both AOs are reacting with free radical during the
isothermal stage of the test, leading to a longer OIT value. This
means the Std-OIT is able to measure both I-168 and I-1010 in the
sample. On the other hand, the testing temperature of HP-OIT is at
150
C which is lower than the effective temperature range of I-168,
but is within the effective range of I-1010. Therefore, the HP-OIT
test is able to measure I-1010, but not I-168. Although Thomas
and Ancelet (1993) were able to obtain an initial HP-OIT value of
129 min for a sample containing 1000 ppm of I-168 and 2.5wt% of
CB and detected a decrease in the HP-OIT value after oven aging at
105
C, they did not clarify whether there was AO presence in the
CB master batch. The 129 min of HP-OIT value could be contributed
by the AOs in the CB package with an effective temperature range
below 150
C.
Adding 2.5 wt% of CB to the AO stabilized HDPE samples, the
initial Std-OIT value did not change significantly as shown in Fig. 1.
This consistent with results obtained by Pena et al., 2001 (b); they
showed no significant Std-OIT change on I-1010 with and without
CB. Contrary, the initial HP-OIT value of samples with CB exhibited a
much lower value than the corresponding non-CB samples, as
shown in Fig. 2. The difference in HP-OIT values between samples
with and without CB is rather significant, ranging from 46 to 62%.
Hawkins et al. (1959a,b) and Gilroy and Chan (1984) found that CB
shortens the induction time using the oxygen absorption test at
temperature from 120 to 140
C which is similar to the isothermal
testing temperature of HP-OIT. One of the possible explanations for
the differ CB effects on the OIT value between Std and HP tests may
be due to the reactivity of AO with CB and with free radicals at
150
C and 200
C. The AO/CB reactivity may be less temperature
dependent than that of the AO/free radical reaction. At 200
C of the
Std-OIT test, the AO/free radical reaction dominated the system and
overmastered the AO/CB interaction. One might suggests that the
low HP-OIT values of CB samples could be caused by adsorption of
AO onto CB surface. However Pena et al., 2001b studied the
adsorption of AOs onto four different types of CB. They found that
the amount of I-1010 being absorbed by the two highly acidic CB
(pH 2.5) is very low, about 0.2 mmol/m2. The adsorption of other
two CB, with pH 8.5, is even lower, about 0.05 mmol/m2. At such low
adsorption level, it is unlikely to lead to significant drop of HP-OIT
value.
4.2. OIT of oven aged samples
To assess the effectiveness of the eight AO formulations and the
interactions between CB and AO, test samples were incubated in a
forced air oven at 85
C to accelerate the oxidation reactions. The

Table 2
Antioxidant content in compression molded plaque and initial OIT value of all
samples.

Sample I-1010 (ppm) I-168 (ppm) Std-OIT HP-OIT

(min.) (min.)
Added Plaque Added Plaque

PE0 0 0 0 0 0.48 na
PE0-C 0 0 0 0 1.2 na
PE1 500 355 0 0 19.4 263
PE1-C 500 368 0 0 21.0 122
PE2 500 199 1000 537 33.3 231
PE2-C 500 342 1000 745 39.1 144
PE3 1000 749 0 0 34.0 339
PE3-C 1000 762 0 0 37.2 189
PE4 1000 836 1000 718 58.7 342
PE4-C 1000 799 1000 714 56.4 194

Note: na e not applicable. Fig. 1. Comparing initial Std-OIT values of the samples with and without CB.
644 W.-K. Wong, Y.G. Hsuan / Geotextiles and Geomembranes 42 (2014) 641e647

Fig. 4. Std-OIT percentage retained versus aging time at 85
C oven incubation.

Fig. 2. Comparing initial HP-OIT values of the samples with and without CB.

AO depletion of the incubated samples was monitored using both

ðOITÞt
Std-OIT and HP-OIT tests. x100 ¼ 100ekt (1)
ðOITÞo
The change of Std-OIT value during oven aging is presented by
plotting Std-OIT value versus incubation time in Fig. 3. For PE1 and
where: (OIT)t ¼ OIT value at different aging time; (OIT)o ¼ initial
PE3 which contain only I-1010 without CB, their Std-OIT values
OIT value; k ¼ reaction rate; and t ¼ aging time;
remain almost constant throughout the 1200 days of incubation.
Fig. 5 shows the HP-OIT plotting against incubation time for the
Interestingly, the Std-OIT value of PE2 decreases steadily with time
four non-CB samples. Samples with similar I-1010 retained con-
until 800 days after which the values are similar to those of PE1 for
centrations gave almost the same HP-OIT value throughout the
the rest of the incubation. The same behavior is also found in PE3
incubation. PE1 and PE2 which contained approximately
and PE4. The converging of the two OIT curves for samples with and
200e350 ppm of I-1010 exhibited an average HP-OIT value of
without I-168 suggests that the decrease of Std-OIT in PE2 and PE4
260 ± 20 min, whereas PE3 and PE4 with 700e830 ppm of I-1010
in the first 800 days of the incubation is contributed to the loss of I-
yielded HP-OIT values at about 330 ± 20 min. The results imply that
168. Fig. 4 plots the Std-OIT retained percentage versus time to
the depletion of I-1010 was minimal during the oven aging period
normalize the different concentrations of I-1010 in the samples.
which was consistent with the Std-OIT data.
Two distinct trends can be seen. Samples PE1 and PE3, which
For the CB samples (PE1-C, PE2-C, PE3-C, and PE4-C), the Std-
contain only I-1010, their Std-OIT values remain at 90% even after
OIT in minute and percentage retained is plotted against incuba-
1200 days of incubation. On the other hand, for PE2 and PE4 which
tion time as can be seen in Fig. 6(a) and (b). The decreasing trend of
contain I-1010 and I-168, their Std-OIT values decrease to about 50%
the four samples can be fitted with Equation (1), and their reaction
retained for the same period of time. The Std-OIT decreasing rate is
rates range from 0.00301 to 0.00391 day1. Compared to samples
independent to the AO concentrations and can be expressed in a
without CB, see Fig. 4, these reaction rates are substantially faster,
first order reaction law, Equation (1), from which the reaction rate
indicating that CB was reacting with one or both types of AO during
of PE2 and PE4 are calculated to be 0.00059 and 0.00056 day1,
the oven aging. Adding 1000 ppm of I-168, the depletion rate
respectively. The dotted line shows the upper and lower bound of
increased slightly from 0.0033 to 0.00391 day1 for PE1C and PE2C;
95% confidence interval which is calculated using student's t-dis-
and from 0.00301 to 0.00343 day1 for PE3-C and PE4C.
tribution method. As for the samples with I-1010 only (PE1 and
For the HP-OIT tests, the HP-OIT value decreased exponentially
PE3), their Std-OIT data cannot be fitted with Equation (1) because
with time for all four CB samples, as shown in Fig. 7(a). In contrast
the OIT value exhibits insignificant changes for the incubation
to the four samples without CB (Fig. 5), which exhibited a merely
period. It is considered that the reaction rate of I-1010 with free
10% decrease in HP-OIT values throughout the incubation, the
radicals in the 85
C oven is negligible.

Fig. 3. Std-OIT value versus aging time at 85
C oven incubation. Fig. 5. HP-OIT value versus time at 85
C oven incubation.
 W.-K. Wong, Y.G. Hsuan / Geotextiles and Geomembranes 42 (2014) 641e647 645

(a) (a)

(b)

(b)

Fig. 6. Std-OIT data of the samples with CB (a) OIT value versus time (b) OIT retained
percentage versus time.

decrease of HP-OIT in CB samples was caused by the interaction

between CB and I-1010. In Fig. 7(b), all four HP-OIT curves collapse
into one when plotting HP-OIT percentage retained versus time.
Fig. 7. HP-OIT data of the samples with CB (a) OIT value versus time (b) OIT retained
Using Equation (1), the reaction rate is calculated to be 0.0010 percentage versus time.
day1 among the four samples.

Std-OIT value was obtained from PE1 and PE2, and PE3 and PE4.
5. Discussions Furthermore, the HP-OIT values remained the same throughout the
aging for samples with same amount of I-1010 with and without I-
The depletion of I-1010 and I-168 in HDPE samples without CB 168. Therefore, the interaction between the two AOs was unlikely.
during oven aging at 85
C was examined using the two OIT tests. Since the first two phenomena were disproved, the possible causes
As shown in Fig. 4, the Std-OIT of PE1 and PE3 remained relatively for the depletion of I-168 may be by reacting with free radicals and
constant throughout the incubation period while exponentially hydrolysis.
decreasing trends in the first 800 days were obtained for PE2 and
PE4. Since PE2 and PE4 contained 1000 ppm of I-168, the decrease
of Std-OIT was likely corresponding to the depletion of I-168. Such
depletion may be due to four phenomena: 1) AO evaporation, 2)
interaction between I-168 and I-1010, 3) reaction between I-168
and free radical, and 4) hydrolysis of I-168. To investigate the effect
of AO evaporation, the OIT profile across the sample thickness was
examined by slicing the samples along the thickness and tested for
their Std-OIT value. Fig. 8 shows the profiles of 0, 850, and 1050 day
of PE2. The initial OIT distribution, at 0 day, indicated an uneven
distribution, 34.5 min of OIT on one side of the surface and rela-
tively constant value across the remaining thickness, ranging from
28.78 to 25.15 min. No significant drop of Std-OIT value at the
surfaces was observed after 1050 days of incubation, implying that
the evaporation of I-168 near the surface of the sample was small.
€rner and Lang (1998) also found that the evaporation
Similarly, Do
of I-168 is negligible for their oven aged samples. For the second
phenomenon, if I-168 reacted with I-1010, the remaining I-1010 in
the samples should be lowered, and thus lower OIT values should
be detected in PE2 and PE4 (with I-168) compared with PE1 and
PE3 (I-1010 only). However, after 1200 days of incubation, a same Fig. 8. Profile of OIT throughout the thickness of PE2 during the incubation.
646 W.-K. Wong, Y.G. Hsuan / Geotextiles and Geomembranes 42 (2014) 641e647

Table 3
Reaction Rate of all eight formulations.

Sample name Formulation Analysis Proposed interaction

AO combination CB (wt.%) Reaction rate (day1) R2 value

PE1 I-1010 (500 ppm) 0 Negligible N/A I-1010/free radical

PE2 I-1010 (500 ppm) þ I-168 (1000 ppm) 0 0.00059 0.937 I-1010/free radical þ hydrolysis of I-168
PE3 I-1010 (1000 ppm) 0 Negligible N/A I-1010/free radical
PE4 I-1010 (1000 ppm) þ I-168 (1000 ppm) 0 0.00056 0.845 I-1010/free radical þ hydrolysis of I-168
PE1-C I-1010 (500 ppm) 2.5 0.00330 0.890 I-1010/free radical þ CB/I-1010
PE2-C I-1010 (500 ppm) þ I-168 (1000 ppm) 2.5 0.00391 0.850 I-1010/free radical þ hydrolysis of I-168
þ CB/I-1010 þ CB/I-168
PE3-C I-1010 (1000 ppm) 2.5 0.00301 0.957 I-1010/free radical þ CB/I-1010
PE4-C I-1010 (1000 ppm) þ I-168 (1000 ppm) 2.5 0.00343 0.929 I-1010/free radical þ hydrolysis of I-168
þ CB/I-1010 þ CB/I-168

I-168 is designed to be a processing stabilizer which interrupts
the chain reaction of generating free radical. Their protective
function of reacting with free radicals at elevated temperature had
been well documented (Kriston et al., 2010, Hinsken et al., 1991,
Schwetlick et al., 1991, Drake et al., 1990, Moss and Zweifel, 1989).
At temperature below 100
C, phosphate could still react with
alkyl-hydroperoxide (ROOH), however, their reactivity was less
effective (Tocha

cek and Sedla
r, 1993). Schwetlick et al., (1988)
conducted series of experiments at 65
C and found that the re-
action rate of peroxyl radicals with phosphites are 10e100 times
lower than the phenols. Although the possibility of the reaction
between phosphites and free radicals cannot be eliminated in this
study, the reactivity appears to be minimal. Lastly, the hydrolytic
stability of phosphites, especially at high temperature (above
100
C), has been studied by many researchers (Kriston et al., 2009,
Neri et al.,1995, Tocha

cek and Sedla
r,1993, Tocha


r,
cek and Sedla
1995, Schwetlick et al., 1991, Holcik et al., 1983). In contrast, the
hydrolytic behavior at temperature below 100
C is less known and
some contradictive results have been reported. Haider and Karlsson
(2002) found the hydrolytic products of I-168 in MDPE film after
exposed at room temperature for 4 years. Tocha

r
cek, and Sedla
(1993) observed about 5% decrease of I-168 after the samples
placed in an environment with water vapor at 25
C for 50 days.
However, Djouani et al. (2011) exposed the I-168 powder to wet
atmosphere for 3.5 months at 80
C and did not detect the product
of hydrolysis. Based on the literature mentioned above, no defini-
tive conclusion can be drawn. It is possible for hydrolysis occurred
during the incubation and causing the depletion of I-168.
For samples with CB, a rapid decrease of OIT was detected in
both Std-OIT and HP-OIT tests. Since the Std-OIT is capable to detect
both AOs, the focus of the analysis is on the Std-OIT data. Table 3
shows the reaction rates and R2) obtained from the oven aged
HDPE test samples together with the proposed interaction(s).
Overall, four interactions are involved in these eight samples, and
they are between CB andI-1010 (CB/I-1010), I-1010 and free radical
(I-1010/free radical), I-168 and moisture (hydrolysis of I-168), and
CB and I-168 (CB/I-168). Assuming the four interactions take place
independently, the total reaction rate can be expressed as the sum
of all four, see Eq. (2). The interaction between CB and I-1010 can be
determined by comparing the reaction rate of PE1-C to PE1 and
PE3-C to PE3. Since the I-1010/free radical reaction rates in PE1 and
PE3 are negligible, the reaction rate of CB/I-1010 should correspond
to those of PE1-C and PE3-C. In contrast, the interaction between CB
and I-168 is not obvious, even though CB/I-168 interaction is one of
the four reactions took place in PE2-C and PE4-C.
1010/free radical, k3 ¼ reaction rate between hydrolysis of I-168,
k4 ¼ reaction rate between CB/I-168.
To determine the reactivity of CB/I-168 (k4), the other three
reaction rates must first be identified. The reaction rate of CB/I-1010
(k1) can be obtained from PE1-C and PE3-C, with values of 0.00330
and 0.00301 day1, respectively. The interaction of I-1010/free
radical (k2) is negligible within the incubation duration. The reac-
tion rates obtained from PE2 and PE4 are mainly due to the hy-
drolytic reaction of I-168, i.e., hydrolysis of I-168 (k3). Since k1, k2
and k3 are now determined, and their summations are 0.00389 and
0.00357 day1 for samples with 500 ppm and 1000 ppm of I-1010,
respectively, the CB/I-168 reaction rate (k4) can be deduced by
subtracting the overall reaction rate obtained from PE2-C and PE4-
C to the sums of k1, k2 and k3. The results are comparatively small,
2  105 day1for PE2-C and no addition effects on PE4-C. Thus it
can be concluded that the reaction between CB and I-168 is
negligible.
6. Conclusions
 The Std-OIT method can detect I-168 and I-1010 and the initial
Std-OIT value is governed by the type and concentration of these
two AOs.
 The HP-OIT test is not sensitive to the presence of I-168. Thus,
the HP-OIT value reflects only the relative amount of I-1010 in
the samples.
 The exponential decrease of I-168 detected by the Std-OIT test
during the oven aging is likely caused by hydrolysis rather than
the oxidation reaction.
 Antagonistic interaction was found between CB and I-1010,
leading a rapid decrease of values in both OIT tests.
 The interactions of CB/I-168 and I-168/free radical during the
oven aging are negligible
Acknowlegement
The project was funded by the Florida Department of Trans-
portation. Sample plaques were supplied by Dr. Shuya Chang of
ExxonMobil Chemical Company. The paper was prepared while Dr.
Hsuan is the Director of the Materials Engineering and Processing
Program in the Civil, Mechanical and Manufacturing Innovation
Division at the National Science Foundation. Any opinions, findings,
and conclusions or recommendations expressed in this material are
those of the authors and do not necessarily reflect the views of the
National Science Foundation.

ktotal ¼ k1 þ k2 þ k3 þ k4 (2)
References
where: ktotal ¼ overall reaction rate of the sample, t ¼ aging time, Djouani, F., Richaud, E., Fayolle, B., Verdu, J., 2011. Modeling of thermal oxidation of
k1 ¼ reaction rate between CB/I-1010, k2 ¼ reaction rate between I- phosphite stabilized polyethylene. Polym. Degrad. Stab. 96, 1349e1360.
 W.-K. Wong, Y.G. Hsuan / Geotextiles and Geomembranes 42 (2014) 641e647
Do€ rner, G., Lang, R.W., 1998. Influence of various stabilizer systems on the ageing
behavior of PE-MD-II. Ageing of pipe specimens in air and water at elevated
temperatures. Polym. Degrad. Stab. 62 (3), 431e440.
Drake, W.O., Pauquet, J.R., Todesco, R.V., Zweifel, H., 1990. Processing stabilization of
polyolefins. Angew. Makromol. Chem. 176 (1), 215e230.
Epacher, E., Fekete, E., Gahleitner, M., Puka
nszky, B., 1999a. Chemical reactions
during the processing of stabilized PE: 1. discolouration and stabilizer con-
sumption. Polym. Degrad. Stab. 63 (3), 489e497.
Epacher, E., Tolve
th, J., Stoll, K., Puka
nszky, B., 1999b. Two-step degradation of high-
density polyethylene during multiple extrusion. J. Appl. Polym. Sci. 74 (6),
1596e1605.
Fay, J.J., King, R.E., 1994. Antioxidants for geosynthetic resins and applications. In:
Hsuan, Y.G., Koerner, R.M. (Eds.), Geosynthetic Resins, Formulations and
Manufacturing, GRI 8 Conference Series. IFAI, St. Paul, Minn., pp. 77e96.
Gilroy, H.M., Chan, M.G., 1984. Effect of pigments on the aging characteristics of
polyolefins. Polym. Sci. Technol. 26, 273.
Haider, N., Karlsson, S., 2002. Loss and transformation products of the aromatic
antioxidants in MDPE film under long-term exposure to biotic and abiotic
conditions. J. Appl. Polym. Sci. 85 (5), 974e988.
Hawkins, W.L., 1984. Polymer Degradation and Stabilization. Springer-Verlag, New
York.
Hawkins, W.L., Hansen, R.H., Matreyek, W., Winslow, F.H., 1959a. The effect of car-
bon black on thermal antioxidants for polyethylene. J. Appl. Polym. Sci. 1 (1),
37e42.
Hawkins, W.L., Matreyek, W., Winslow, F.H., 1959b. The morphology of semi-
crystalline polymers. Part I. The effect of temperature on the oxidation of
polyolefins. J. Polym. Sci. 41 (138), 1e11.
Hinsken, H., Moss, S., Pauquet, J., Zweifel, H., 1991. Degradation of polyolefins during
melt processing. Polym. Degrad. Stab. 34 (1e3), 279e293.
Holcik, J., Koenig, J., Shelton, J.R., 1983. The antioxidant activity of phosphorus
compounds e part I: decomposition of hydroperoxides by pentaerythritol
diphosphites. Polym. Degrad. Stab. 5, 373e397.
Hsuan, Y.G., Koerner, R.M., 1998. “Antioxidant depletion lifetime in high density
polyethylene geomembranes”. J. Geotechn. Geoenviron. Eng. 124, 532.
Kovacs, E., Wolkober, Z., 1976. The effect of the chemical and physical properties of
carbon black on the thermal and photooxidation of polyethylene. J. Polym. Sci.
Symposium No. 57 p.171.
Kriston, I., Orba
n-Mester, A.,
Nagy, G., Staniek, P., Fo €ldes, E., Puka
nszky, B., 2009.
Melt stabilisation of Phillips type polyethylene, Part II: correlation between
additive consumption and polymer properties. Polym. Degrad. Stab. 94,
1448e1456.
Kriston, I., Pe
nzes, G., Szijja
rto
, G., Szabo

, P., Staniek, P., Fo
€ ldes, E., Puk

anszky, B.,
2010. Study of the high temperature reactions of hindered aryl phosphite
(Hostanox PAR 24) used as a processing stabilizer in polyolefin. Polym. Degrad.
Stab. 95, 1883e1893.
Lundb€ ack, M., Strandberg, C., Albertsson, A., Hedenqvist, M.S., Gedde, U.W., 2006.
Loss of stability by migration and chemical reaction of Santonox® R in branched
polyethylene under anaerobic and aerobic conditions. Polym. Degrad. Stab. 91,
1071e1078.
Malle
gol, J., Carlsson, D.J., Desche ^nes, L., 2001. A comparison of phenolic anti-
oxidant performance in HDPE at 32e80
C. Polym. Degrad. Stab. 73 (2),
259e267.
Moore, L.M., Marshall, G.P., Allen, N.S., 1988. Degradation and stabilisation of blue
water pipe. Polym. Degrad. Stab. 20 (3e4), 337e354.
647
Moss, S., Zweifel, H., 1989. Degradation and stabilization of high density poly-
ethylene during multiple extrusions. Polym. Degrad. Stab. 25 (2e4), 217e245.
Muller, W., Jacob, I., 2003. Oxidative resistance of high density polyethylene geo-
membranes. Polym. Degrad. Stab. 79 (No. 1), 161e172.
Nagy, K., Epacher, E., Staniek, P., Puka
nszky, B., 2003. Hydrolytic stability of phenolic
antioxidants and its effect on their performance in high-density polyethylene.
Polym. Degrad. Stab. 82 (2), 211e219.
Neri, C., Costanzi, R.M., Riva, R., Farris, R., Colombo, R., 1995. Mechanism of action of
phosphites in polyolefin stabilisation. Polym. Degrad. Stab. 49, 65e69.
Pen~ a, J.M., Allen, N.S., Edge, M., Liauw, C.M., Valange, B., 2001a. Factors affecting the
adsorption of stabilisers on to carbon black (flow micro-calorimetry studies) 4.
Secondary antioxidants. Polym. Degrad. Stab. 72 (1), 31e45.
Pen~ a, J.M., Allen, N.S., Edge, M., Liauw, C.M., Valange, B., 2001b. Interactions be-
tween carbon black and stabilisers in LDPE thermal oxidation. Polym. Degrad.
Stab. 72 (1), 163e174.
Phease, T.L., Billingham, N.C., Bigger, S.W., 2000. The effect of carbon black on the
oxidative induction time of medium-density polyethylene. Polymer 41 (26),
9123e9130.
Pleshanov, V.P., Berlyant, S.M., Burukhina, G.A., 1982. Study of kinetics of inhibition
of thermo-oxidation of polyethylene by carbon black. Polym. Sci. U.S.S.R. 24 (6),
1462e1468.
Pock, E., Kiss, C., Janecska, A.,
Epacher, E., Puk
anszky, B., 2004. Effect of chain
structure on the processing stability of high-density polyethylene. Polym.
Degrad. Stab. 85 (3), 1015e1021.
Rowe, R.K., Rimal, S., 2008. Depletion of antioxidants from an HDPE geomembrane
in a composite liner. ASCE J. Geotechnical. Geoenviron. Eng. 134 (No. 1), 68e78.
Schwetlick, K., Ko €nig, T., Pionteck, J., Sasse, D., Habicher, W.D., 1988. Organophos-
phorus antioxidants: Part IX e inhibition of the oxidation of hydrocarbons by
hindered aryl phosphites. Polym. Degrad. Stab. 22, 357e373.
Schwetlick, K., Pionteck, J., Winkler, A., H€ ahner, U., Kroschwitz, H., Habicher, W.D.,
1991. Organophosphorus antioxidants: Part X e mechanism od antioxidant
action of aryl phosphites and phosphonites at high temperatures. Polym.
Degrad. Stab. 31, 219e228.
Smith, G.D., Karlsson, K., Gedde, U.W., 1992. Modeling of antioxidant loss from
polyolefins in hot-water applications. I: model and application to medium
density polyethylene pipes. Polym. Eng. Sci. 32, 658e667.
Thomas, R.W., Ancelet, C.R., 1993. The Effect of Temperature, Pressure, and Oven
Aging on the High Pressure Oxidative Induction Time of Different Types of
Stabilizers. Geosynthetics '93. IFAI, Vancouver, Canada, pp. 915e925.
Tocha
cek, J., Sedla
r, J., 1993. Effect of hydrolysability and structural features of
phosphites on processing stability of isotactic polypropylene. Polym. Degrad.
Stab. 41, 177e184.
Tocha

r, J., 1995. Hydrolysis and stabilization Performance of bis(2,4-di-
cek, J., Sedla
t-butylphenyl) pentaerythrityl diphosphite in Polypropylene. Polym. Degrad.
Stab. 50, 345e352.
Viebke, J., Gedde, U.W., 1997. Antioxidant diffusion in polyethylene hot-water pipes.
Polym. Eng. Sci. 37 (5), 896e911.
Viebke, J., Gedde, U.W., 1998. Assessment of lifetime of hot-water polyethylene
pipes based on oxidation induction time data. Polym. Eng. Sci. 38 (8),
1244e1250.
Wong, W.K., Hsuan, Y.G., 2009. Carbon Black and Antioxidant Effect in High Density
Polyethylene. Proceeding of Geosynthetic 2009, Salt Lake City, Utah.
Wong, W.K., Hsuan, Y.G., 2012. Interaction between carbon black and antioxidant in
HDPE pipe resin. J. Transp. Res. Board 2310, 137e144.