See

discussions, stats, and author profiles for this publication at:

https://www.researchgate.net/publication/278087800

Alkyd resins derived from glycolized waste

poly(ethylene terephthalate)

Article in European Polymer Journal February 2005

DOI: 10.1016/j.europolymi.2004.10.001

CITATIONS READS

29 12

4 authors, including:

Dimitris S. Achilias IRINI Sideridou

Aristotle University of Thessaloniki Aristotle University of Thessaloniki
149 PUBLICATIONS 3,621 CITATIONS 112 PUBLICATIONS 2,221 CITATIONS

SEE PROFILE SEE PROFILE

Dimitrios N Bikiaris
Aristotle University of Thessaloniki
329 PUBLICATIONS 8,793 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Various parameters affecting the fraction and dynamics of polymer in the interfacial layer
with nanoparticles View project

All content following this page was uploaded by IRINI Sideridou on 07 March 2016.

The user has requested enhancement of the downloaded file.

 EUROPEAN
POLYMER
European Polymer Journal 41 (2005) 201210
JOURNAL
www.elsevier.com/locate/europolj

Alkyd resins derived from glycolized waste

poly(ethylene terephthalate)
George P. Karayannidis *, Dimitris S. Achilias,
Irini D. Sideridou, Dimitris N. Bikiaris
Laboratory of Organic Chemical Technology, Department of Chemistry, Aristotle University of Thessaloniki, GR-541 24,
Thessaloniki, Macedonia, Greece

Received 3 September 2004; received in revised form 28 September 2004; accepted 1 October 2004

Abstract

In this investigation the production of secondary value-added products, such as alkyd resins, derived from the gly-
colysis of poly(ethylene terephthalate) (PET) is examined as an eective way for its recycling. PET was taken from com-
mon soft-drink bottles and diethylene glycol (DEG) was used for the depolymerization at several initial molar ratios.
The oligomers obtained were analyzed according to their average molecular weights. Furthermore, the glycolyzed PET
products (oligomers) were reacted with maleic anhydride, phthalic anhydride and propylene glycol to form unsaturated
polyester resins. These were subsequently mixed with styrene and cured using the benzoyl peroxide/amine initiator sys-
tem to carry out the reaction in ambient temperature. The curing characteristics of the resins produced were investi-
gated with respect to the initial molar ratio of DEG/PET as well as the initial initiator concentration. Finally, the
mechanical properties (tensile strength and elongation at the break point) of the resins were compared with the conven-
tional general purpose resin and were found to be comparable.
2004 Elsevier Ltd. All rights reserved.

Keywords: Recycling; Alkyd resins; Curing of unsaturated polyesters; Glycolysis; Poly(ethylene terephthalate)

1. Introduction rial are consumed in the manufacture of video and audio

tapes, X-ray lms, food packaging and especially of soft-
Poly(ethylene terephthalate) (PET) is a thermoplastic drink bottles.
polyester showing excellent thermal and mechanical PET recycling represents one of the most successful
properties. Although its main application was by far and widespread examples of polymer recycling. PET
the textile industry, tremendous quantities of this mate- bottle collection in Europe (European Union member
states plus Norway, Iceland, Switzerland, and all EU
candidate countries) is growing steadily. In 2001
*
Corresponding author. Tel.: +30 2310 997814; fax: +30 344,000 ton were collected, a 20% increase in comparison
2310 997769. with 2000. By 2006, it is forecast that European PET col-
E-mail address: karayan@chem.auth.gr (G.P. lection will increase to 700,000 ton [1]. The main driving
Karayannidis). force responsible for this extremely increased recycling

0014-3057/$ - see front matter 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.eurpolymj.2004.10.001
202 G.P. Karayannidis et al. / European Polymer Journal 41 (2005) 201210

of post-consumer PET is its widespread use, particularly gated by Aslan et al. [15]. In the same process of glass-
in the beverage industry. A very important feature of ber-reinforced plastics manufacturing, Radenkov et al.
PET, decisive in the choice of its wide application in [16] investigated the possibility of direct using products
the manufacture of packaging for the food industry is from PET glycolysis. Glycolysis of recycled PET was also
that it does not have any side eects on the human investigated by Chen et al. [17] and Mansour and Ikla-
organism. It should be pointed out, that PET does not dius [18]. Finally, the mechanical properties and the
create a direct hazard to the environment but, due to characteristics of UPR produced from post-consumer
its substantial fraction by volume in the waste stream PET bottles was investigated by Pimpan et al. [19] and
and its high resistance to the atmospheric and biological the synthesis of coating resins from recycled PET was
agents, it is seen as a noxious material [2]. Therefore, the investigated recently by Kawamura et al. [20].
recycling of PET does not only serve as a partial solution The aim of this research was to use the unsaturated
to the solid waste problem but also contribute to the oligoester diols produced from PET recycling as poten-
conservation of raw petrochemical products and energy. tial raw materials for the production of alkyd resins used
Among the dierent recycling techniques, the only as enamel paints or coatings. In order to fulll this goal
one acceptable according to the principles of Sustain- three steps were distinguished. The rst step includes the
able Development is the chemical recycling, since it depolymerization of PET using various diethylene gly-
leads to the formation of the raw materials from which col/PET ratios and the identication of the oligomers
the polymer is made of, as well as of other secondary produced. In the second step the curing behavior of
value-added products [3]. A successful recycling pro- unsaturated polyester resins used for the production of
gram does not only depend on post-consumer waste col- alkyd resins for ambient temperature coatings was inves-
lection, it also depends on whether the products made tigated by manipulating the amounts of the initiators
out of collected, reclaimed and recycled material re- used. In the nal step, the glycolized products (glycolyz-
spond to consumers
needs, in other words if recycled ates) taken from PET recycling replace the diols used
products are actually bought. This is the reason why for the conventional production of the alkyd resins
the PET industry constantly researches for reclaimed and the nal mechanical properties of the resins were
material new applications. Recently, a growing interest compared.
has been observed in the use of PET wastes for the pro-
duction of specialized products such as unsaturated
polyester, polyurethane foams and polymer concrete. 2. Experimental
In our previous publications [4,5], the alkaline and
acid hydrolysis of PET was examined in order to take 2.1. Materials
the pure monomer (terephthalic acid) from which the
polymer is made of. This monomer was used as raw PET akes were prepared from post-consumer clear
material for the reproduction of PET. As a step further, PET bottles, from which the polyethylene caps and the
in this paper, glycolysis of PET is examined as a process polypropylene label had been removed. The bottles were
for the production of raw materials used in coating res- cut and fed to a rotary cutter with a max size of 6 mm.
ins. Vaidya and Nadkarni were among the rst who sys- Diethylene glycol (DEG), 1,2-propylene glycol (PG),
tematically studied the formation of unsaturated maleic anhydride (MA) and Styrene monomer were sup-
polyester resins (UPR) from PET glycolysis [6,7]. Baliga plied by Fluka AG. Phthalic anhydride (PA) was
and Wong [8] carried out the glycolysis of PET with eth- obtained from Merck. Manganese acetate [Mn(CH3-
ylene glycol (EG) using various catalysts. They found COO)24H2O] with a purity of 99% was supplied from
that glycolyzed products had 13 repeating units Aldrich. The initiators for the curing reaction of the
depending on the catalyst used. Halacheva and synthesized unsaturated polyester resins were benzoyl
Novakov [9] have investigated the chemical structure peroxide (BPO) from Fluka AG and N,N-dimethyl-
of the oligoesters produced from PET glycolysis with p-toluidine (DMT) from Aldrich. Tetrahydrofuran
diethylene glycol (DEG). Furthermore, Suh et al. [10] (THF) used as a solvent in gel permeation chromatogra-
examined the properties of unsaturated polyesters pro- phy (GPC) was obtained from Riedel-de-Haen. All
duced from glycolyzed PET with various combinations other chemicals used were reagent grade.
of the glycols DEG/propylene glycol (PG). Farahat
and Nikles [11,12] provided results on the glycolysis of 2.2. Glycolysis of PET
PET and the production of novel binder systems for sol-
ventless magnetic tape manufacturing. Farahat et al. In the glycolysis of PET, 70 g (0.365 mol/repeating
also synthesized modied UPR from PET glycolysis unit) of PET waste akes were added into the reactor to-
using dierent DEG/PET molar ratios [13,14]. The pro- gether with the diethylene glycol in three molar ratios
duction of ber-reinforced thermosetting composites [10]. The molar ratios of DEG/PET were 1.20 or
using the product of PET glycolysis has been investi- 46.4 g, 1.60 or 61.9 g and 2.20 or 85.1 g of DEG for
 G.P. Karayannidis et al. / European Polymer Journal 41 (2005) 201210 203

experiments number 1, 2 and 3, respectively. Manganese dx 1 dDH

acetate, 0.5 wt.% based on weight of PET, was added as dt DH T dt
transesterication catalyst. The reactor was immersed in
where DHT denotes the total reaction enthalpy.
an oil bath and heated for 4 h at 210 C. The reaction
was carried out under reux in argon atmosphere. After The polymerization enthalpy and conversion was cal-
the 4 h period, the reactor was allowed to cool at room culated by integrating the area between the DSC ther-
temperature and the reaction product was ltered mograms and the baseline established by extrapolation
through a G3 glass lter. The ltrate was extracted with from the trace produced after complete polymerization
dichloromethane and the extract was nally washed with (no change in the heat produced during the reaction).
water in order to remove traces of DEG. The total reaction enthalpy was determined by heating
the sample from the polymerization temperature to
2.3. Synthesis of unsaturated polyester resin 180 C at a rate of 10 C/min. The sum of enthalpies
of the isothermal plus the dynamic experiment was the
Initially, a reference resin (blank) was synthesized total reaction enthalpy (approximately 300 J/g).
using the following procedure. Into a three-necked resin
kettle equipped with a mechanical stirrer, inlet to inert 2.5. Measurements
gas (argon), thermometer, condenser and calibrated dis-
tillate collector, 76.1 g (0.5 mol) of phthalic anhydride, The molecular weight distribution and the average
49.0 g (0.5 mol) of maleic anhydride, 41.0 g (0.43 mol) molecular weights of the glycolyzed PET products were
of diethylene glycol and 44.1 g (0.58 mol) of propylene determined by gel permeation chromatography. The
glycol were introduced [21]. The reactor was purged with instrument used was from Polymer Laboratories and in-
argon for 15 min before heating was switched on. Then cluded a pump (Marathon III HPLC pump), an Evapo-
the content was heated under argon atmosphere to rative Mass Detector (PL-EMD 950) and a Plgel 5l
200 C until approximately 15 ml of water were collected. MIXED-D column. All samples were dissolved in
This took approximately 9 h. The mixture was then THF at a constant concentration of 0.2 wt.%. After l-
cooled down to 90100 C and diluted with 125 ml of sty- tration of samples, 25 ll of each sample was injected into
rene. The amount of styrene was equal to 0.54 mol/mol the chromatograph. The elution solvent was also THF
of total initially added reactants (PA, MA, DEG and at a constant ow rate of 1 ml/min. Calibration of
PG), or 35 wt.% 15 mg of hydroquinone were also added GPC was carried out with standard polystyrene samples
in order to inhibit initiation of the polymerization. (Polymer Laboratories) by using the universal calibra-
The above experiment was repeated but using the gly- tion technique.
colyzates instead of diethylene glycol and exactly the Tensile tests were performed using an Instron 1122
same molar ratio for the rest of the reactants (i.e. PA, dynamometer, in accordance to the ASTM D638 method
MA and PG). The amount of styrene added nally at room temperature. Prior to measurements the samples
was also kept at exactly the same molar ratio as with were conditioned at 50 5% relative humidity for 36 h,
the reference UPR. placed in a closed chamber containing a saturated
Ca(NO3)2 4H2O solution in distilled water (ASTM E-
2.4. Curing of the unsaturated polyester 104). Tensile tests were performed under the condition
that the crosshead speed of tensile tester was set at
In order to cure the resin in ambient temperature the 5 mm/min and the initial gauge length was xed at
redox system of peroxideamine was used. Benzoyl per- 22 mm. Five measurements were conducted for each sam-
oxide (BPO) and N,N-dimethyl-p-toluidine were intro- ple, and the results were averaged to obtain a mean value.
duced into the mixture as the two coinitiators. The The specimens for the mechanical properties were
kinetics of the free-radical polymerization of the unsatu- prepared by mixing the resin with the two coinitiators
rated resin were investigated using the DSC-Pyris 1 and immediately spreading the product on a glass sur-
(Perkin-Elmer) equipped with the Pyris software for face. Thin lms were thus prepared from curing at room
windows. Indium was used for the enthalpy and temper- temperature.
ature calibration of the instrument. The molar ratio The content of hydroxy end-groups (hydroxyl num-
BPO:amine was kept constant at 1:1 and several initial ber, HN) of the oligoester diols obtained was determined
BPO/resin weight percent were examined. Polymeriza- by titration with 1 N KOH solution. About 1.6 g sample
tions were carried out at 23 C. The reaction exotherm was accurately weighted and added to 10 ml acetylating
(in normalized values, W/g) at a constant temperature solution, containing 10 vol.% acetic anhydride in pyri-
was recorded as a function of time. The rate of heat re- dine, in a round-bottomed ask. The ask was tted
lease (d(DH)/dt) measured by the DSC was directly con- with a vertical reux condenser and heated to 80 C
verted into the overall reaction rate (dx/dt) using the for 2 h. Afterwards 10 ml of water was added and the
following formula: heating was continued for 5 min at 100 C. The mixture
204 G.P. Karayannidis et al. / European Polymer Journal 41 (2005) 201210

was then cooled at room temperature and 10 ml of meth- ucts, which in the next step of the polyesterication reac-
anol were added. The resulting solution was titrated tion produce two-phase unsaturated polyesters. One of
against 1 N KOH standard using phenolphthalein as these phases is incompatible with styrene, which is used
indicator. A blank run was also performed. The values in the curing reaction [15].
obtained provide a means of the number average molec- The glycolysis of PET in this study was carried out
ular weight. using three glycol compositions. The initial molar ratios
of DEG/PET used were 1.20, 1.60 and 2.20 correspond-
ing to glycolyzed products with sample codes GLY-1,
3. Results and discussion GLY-2 and GLY-3, respectively. Results on the average
molecular weights of the three glycolyzates are presented
3.1. Glycolysis of PET in Table 1. In this table results are included on the num-
ber average molecular weight M n of the polyester olig-
The alcoholysis of PET with diethylene glycol pro- omers produced using the hydroxyl number or the GPC.
ceeds according to the following reaction. Also, the weight average molecular weight M w and the

O O

C CO CH2 CH2 O

+
HOCH2 CH2 OCH2 CH2 OH

O O

C CO CH2 CH2 OCH2 CH2 OH

HOCH2 CH2 O

The alcoholysis consists of the transesterication of polydispersity of the molecular weight distribution of
PET and the destruction of its polymer chain, resulting these samples obtained from GPC are included. As the
in the decrease of its molecular weight. Using glycols amount of DEG is increased the depolymerization is
in the depolymerization of PET, the oligoesters obtained carried out to a greater extent leading to oligoester diols
have two hydroxyl end groups, i.e. oligoester diols are with lower number average molecular weights. The same
formed. The choice of DEG to carry out the glycolysis trend in the results observed by either end group analysis
is usually determined by the necessity of having well ex- or GPC, while the results obtained from GPC where al-
ural properties in the UPR, since the long chains of the ways higher than the corresponding from the hydroxyl
glycol improves exibility. Use of other glycols such as number. The value of 454 measured for GLY-3 is very
propylene glycol or dipropylene glycol results in prod- close to the value of 478 measured by Farahat and Ni-

Table 1
Average molecular weights of the glycolized PET products obtained from hydroxyl number and GPC
Glycolized product Molar ratio (DEG/PET) Hydroxyl number (HN) M n from hydroxyl number Data from gel permeation
chromatography
Mn Mw Polydispersity
GLY-1 1.20 189 593 510 602 1.18
GLY-2 1.60 212 528 493 582 1.18
GLY-3 2.20 239 469 454 518 1.14
 G.P. Karayannidis et al. / European Polymer Journal 41 (2005) 201210 205

100 of the components of the glycolyzed PET products could

GLY-1 be done for example by using 13C NMR, or 1H NMR.
GLY-2
80
GLY-3 3.2. Synthesis of unsaturated polyester resin

Initially, a blank experiment was carried out, using

60
the reactants that are usually used in the synthesis of
dw/dlogM

alkyd resins. Afterwards the same experiment was re-

40 peated but using the oligoester diols obtained from
PET glycolysis in place of diethylene glycol. Thus, it is
possible to directly compare the results obtained when
20 using the glycolyzed PET products instead of the con-
ventional raw materials. The basic chemical reaction
occurring in the production of alkyd resins from maleic
0
2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4
anhydride, phthalic anhydride, diethylene glycol and
propylene glycol is shown in Scheme 1. When the glyco-
log M
lyzates are used, then DEG is completely replaced by the
Fig. 1. Molecular weight distribution of the glycolized products oligoester diols. It is obvious that the molecular weight
with various DEG/PET molar fractions. of the structural unit of the unsaturated polyester resin
produced when using the glycolyzates is larger compared
to the blank experiment. For this reason these products
kles [11] using GPC and a DEG:PET molar ratio of had always higher viscosity compared to the blank resin.
2.15:1. The polydispersity of the molecular weight distri- The sample codes used for the UPR produced from the
bution, which was always greater than 1, revealed that glycolyzed products were UPR-1, UPR-2 and UPR-3 in
the distribution was not unimodal. Actually, in every relation to the glycolyzed diols GLY-1, GLY-2 and
sample three distinct peaks (trimodal shape) were identi- GLY-3, used respectively.
ed. The full molecular weight distribution of the glyco-
lyzed PET products obtained at the three dierent initial 3.3. Curing of the unsaturated polyester
molar fractions appears in Fig. 1. The three peaks iden-
tied correspond to molecular weights of 350, 540 and The unsaturated polyesters are long chain polymers
725. That means that PET glycolysis using DEG results containing a number of reactive double bonds. In order
mainly in three kinds of oligoesters. As the amount of to form crosslinked networks, they are dissolved in a
DEG was increased the area under the peak correspond- free-radical polymerizable monomer such as styrene.
ing to the lowest molecular weight was increased and This polymerizable monomer, which also contains reac-
respectively the area under the peak of the higher molec- tive double bonds (C@C), acts as a curing agent by
ular weight was lowered. Hence, with increased amount bridging adjacent polyester molecules at their unsatura-
of DEG a better depolymerization of PET was obtained. tion positions. The content of styrene in the nal resin is
On trying to identify the oligoester diols produced important to ensure good processability. Moreover, if its
the following structure was proposed [9]: content is very high or very low, we can have high

O O O O

HOCH2 CH2 OCH2 CH2 O C COCH2 CH2 O C COCH2 CH2 OCH2 CH2 OH

where m = 0
3. shrinkage on curing or a tendency not to cure fully,

According to the above structure if one replaces m respectively. Vaidya and Nadkarni [6] proposed an eec-
with 0, 1 and 2 the following molecular weights are ob- tive range for the styrene percentage in a polyester resin
tained: 342, 534 and 726, respectively. By comparing from 30 to 40 wt.%. In this investigation a 35 wt.% for
these values with the values obtained by GPC, it can the blank resin was used, in accordance also to other lit-
be concluded that the above structure is conrmed with erature sources [10]. The curing behavior of unsaturated
m = 0, 1 and 2. Similar structures have been observed polyester resins has been extensively studied in literature
also in literature [11]. However, a denite identication [2228]. The polymerization is usually initiated using a
206 G.P. Karayannidis et al. / European Polymer Journal 41 (2005) 201210

O O

C C

n O + n O + n H A H + n HO CH2CH OH

C C CH3
O O

O O O O

A C C OCH2CHO C C + 2n H2O
HC CH
CH3

where A in the case of the blank experiment is OCH2CH2 OCH 2CH 2O

and in the case of the oligoester diols it takes the following form (m=0,1,2) :

O O O O

OCH2CH2OCH2CH2O C COCH2CH2O C COCH2CH2OCH2CH2O

Scheme 1. Synthesis of unsaturated polyester resins using either diethylene glycol, or the oligoesters diols derived from PET glycolysis.
It should be noted that the places of the two diols in the UPR structural unit could be interchanged. The same holds for the place of the
two anhydrites.

free radical initiator such as benzoyl peroxide. However, As it was expected, an increase in the initial initiator
it requires elevated temperatures (5090 C) in order to concentration leads to higher reaction rates accompa-
decompose the initiator molecule and form the primary nied by shorter induction times, higher maximum poly-
radicals needed. If the curing must be carried out in merization rate and completion of the reaction at
ambient temperature, then a peroxide/amine system is shorter times. According to the results of Fig. 2, a weight
used to form the primary free radicals [21]. As such, fraction of BPO 1% may be adequate in industrial use,
we used the benzoyl peroxide/N,N-dimethyl-p-toluidine, when the curing must be completed fast. However, in
already studied for the free-radical polymerization of laboratory scale experiments the low weight fraction of
MMA [29]. To prevent premature gelation and to inhibit 0.5% is preferable in order to have enough time to pro-
the spontaneous polymerization of the resin, hydroqui- cess the material and produce lms that are necessary
none was added as an inhibitor [15]. for preparing the specimens that will be used in the
The free-radical polymerization reaction and the for- mechanical properties testing.
mation of the crosslinked network after the curing of the The total heat of polymerization (i.e. the sum of the
UPR with styrene is shown in Scheme 2. isothermal and the dynamic period) was approximately
In order to nd the best initial weight fractions of the equal to 300 J/g, which is very close to corresponding lit-
initiators, so that the polymerization to complete in a erature values [15,22,25,27]. The ultimate total double
short time period, while long enough to be detected by bond conversion was not inuenced much by the initial
DSC, several tests were carried out. Results on the time initiator concentration and it was approximately equal
dependence of the reaction rate and double bond con- to 68%.
version at three initial initiator concentrations, appear Furthermore, the curing rate characteristics of the
in Fig. 2a and b. It should be pointed out that always UPR-1, UPR-2 and UPR-3 alkyd resins produced from
the amine and the peroxide were in equimolar amounts. the glycolyzed PET products are compared to the blank
 G.P. Karayannidis et al. / European Polymer Journal 41 (2005) 201210 207

O O O O

A C C OCH2CHO C C + m HC CH2
HC CH
CH3

BPO
23oC
DMT

O O O O O O O O
A C C OCH2CHO C C A C C OCH2CHO C C
HC CH CH CH
CH3
CH3
CH
CH2 CH2
CH2
HC HC

CH2 CH CH O O

CH A C C OCH2CHO C C
O O CH3
CH2
CH CH
CH
CH2 CH2
CH CH O O CH CH O O
A C C OCH2CHO C C A C C OCH2CHO C C

O O CH3 O O CH3

Scheme 2. Curing reaction of the UPR with styrene.

experiment in Figs. 3 and 4 for two dierent initial initi- with UPR-2 compared to blank and UPR-1. In both ini-
ator concentrations 1 wt.% and 0.7 wt.%, respectively. A tial initiator concentrations the start of the reaction was
marked increase in the reaction rate is observed initially, immediate without any induction time, the maximum in
which is caused by the strong inuence of diusional the curing rate occurred earlier and the peak was more
phenomena on the termination rate constant, together pronounced. Compared to the blank resin, the greater
with the formation of the polymer network. At the peak reactivity of the UPR-2 can be attributed to a reduced
of the reaction rate the network has become quite dense steric hindrance of the reactive double bonds due to
and thus the diusion of small molecules, such as mono- the longer chain of the repeated structural unit. UPR-1
mers, is hindered leading to a reduction in the reaction with a slightly higher average chain length compared
rate until it reaches a zero value. In every case the reac- to UPR-2 exhibits a slightly slower reaction rate. How-
tion ceases when the glass transition temperature of the ever, this trend was not followed by UPR-3 since this
reacting mixture equals to the polymerization tempera- resin exhibited a very high viscosity (due to a lower
ture, though the monomers have not reacted completely. amount of styrene added as a diluent). Hence the maxi-
On comparing the reaction rate characteristics between mum in the reaction rate was lowered compared to
the dierent resins obtained from the PET glycolysis UPR-2 and the reaction lasted more than all the other
diols to the blank experiment it can be noticed that in cases. The maximum double bond conversion obtained
both initial initiator concentrations UPR-1 exhibits al- with UPR-1 was around 70%, slightly higher than the
most the same characteristics as the blank resin. The corresponding of the blank resin.
induction period is almost the same, as well as the value
of the maximum in the reaction rate and the time at 3.4. Mechanical properties of the alkyd resins
which it appears. A slight decrease in the induction time
was observed at the initial BPO concentration of 1 wt.%. Finally, the mechanical properties of the alkyd resins
However, a much faster reaction rate has been observed produced were studied in relation to the eect of the
208 G.P. Karayannidis et al. / European Polymer Journal 41 (2005) 201210

-3 -3
1.2x10 1.0x10
BPO 0.5% wt -4
9.0x10 Blank
1.0x10
-3 BPO 0.7% wt
-4
8.0x10
UPR-1
BPO 1.0% wt
UPR-2
-4
-4 7.0x10 UPR-3
8.0x10
Curing Rate (1/s)

Curing Rate (1/s)

-4
6.0x10
-4
6.0x10 -4
5.0x10
-4
4.0x10
-4
4.0x10
-4
3.0x10
-4 -4
2.0x10 2.0x10
-4
1.0x10
0.0
0.0
0 50 100 150 200
0 20 40 60 80 100 120
(a) Time (min)
Time (min)
100
Fig. 4. Rate of curing as a function of time for the UPR
BPO 0.5 % wt produced from the glycolyzed oligoester diols and the blank
BPO 0.7 % wt resin. Initial BPO weight fraction 0.7%.
80
BPO 1.0 % wt
Double bond conversion (%)

initial initiator concentration, as well as to the eect of

60 the oligoester diols used. In all resin samples (containing
MA, PA, PG and either DEG or the glycolyzates to-
gether with styrene) the mixture of the two initiators
40
was added and thin lms were prepared that were mea-
sured at dierent curing times from 1 day to 40 days.
20 The eect of the curing time and of the initial initiator
concentration on the elongation at the break point and
the tensile strength of the blank resin is shown in Fig.
0
0 50 100 150 200 5a and b, respectively. As it can be seen the mechanical
(b) Time (min) properties of the resins are time dependent and they
seem to reach a steady value after approximately 20 days
Fig. 2. Curing rate (a) and double bond conversion (b) versus of curing period. Low curing rates cause a brittleness in
time of the blank resin using dierent initial initiator (BPO) the resins, which results in its low tensile strength. An in-
weight fractions. The amine/BPO molar fraction was always
crease in the initial initiator concentration leads to high-
1/1.
er tensile strength and lower elongation at break. It
seems that using a high initial initiator concentration
the crosslinked network formed is denser. Thus, the
-3
1.4x10 chains are not able to move freely which makes the resin
Blank
-3 UPR-1 less exible. Furthermore, at the same initial initiator
1.2x10
UPR-2 concentration the eect of the resins produced from
UPR-3 the oligoester diols on the tensile strength and elonga-
Curing Rate (1/sec)

-3
1.0x10
tion appear in Fig. 6a and b, respectively. Both alkyd
8.0x10
-4
resins produced from the glycolyzed PET products exhi-
bit lower tensile strength and higher elongation at the
-4
6.0x10 break point, compared to the blank resin. This must
-4
be attributed to the longer distance between the reactive
4.0x10
double bonds existing in the UPR derived from the gly-
2.0x10
-4 colyzed products compared to that of the blank resin.
This enables the cured resin to have a lower crosslinking
0.0 density and thus to become more exible. Comparing
0 10 20 30 40 50 60 70 80
the resins UPR-1 and UPR-2, the tendency for the ten-
Time (min)
sile strength is to become lower and for the elongation to
Fig. 3. Rate of curing as a function of time for the UPR become higher when the resin used has lower molecular
produced from the glycolyzed oligoester diols and the blank weight. The reason to that can be attributed to the eect
resin. Initial BPO weight fraction 1%. of the molecular weight of the repeating unit dierence
 G.P. Karayannidis et al. / European Polymer Journal 41 (2005) 201210 209

20 50
BPO 1.0%-wt Blank
18 45
BPO 0.75%-wt UPR-1
16 BPO 0.5%-wt 40 UPR-2
14 35
Elongation (%)

Elongation (%)
12 30

10 25

8 20

6 15

4 10

2 5

0 0
0 10 20 30 40 0 10 20 30 40
(a) Curing time (days) (a) Curing time (days)
55
50
50
45
45
40
40
Tensile strength (MPa)

Tensile strength (MPa)

35
35
30
30

25 25

20 20

15 15

10
BPO 1.0%-wt
BPO 0.75%-wt
10 Blank
5 BPO 0.5%-wt UPR-1
5
UPR-2
0
40 0
0 10 20 30
0 10 20 30 40
(b) Curing time (days) (b) Curing time (days)
Fig. 5. Eect of the curing time on the elongation at the break
Fig. 6. Eect of the curing time on the elongation at the break
point (a) and the tensile strength (b) of the blank resin cured
point (a) and the tensile strength (b) of the blank resin and the
using dierent initial initiator concentrations.
resins produced for the glycolyzed PET diols cured at 0.5 wt.%
initial BPO concentration.
between resins UPR-1 and UPR-2. The former exhibits
on average a higher molecular weight and thus it con-
tains more terephthalate repeating units. The presence gated. The oligoester diols produced from the reaction
of more terephthalate repeating units between the cross- of DEG with PET consisted mainly of three oligomers.
links results in the existence of harder domains and bet- An increase in the molar ratio of DEG:PET led to better
ter separation between the crosslinks in the packed depolymerization and to an increase in the amount of
crosslinked structure and as a result the mechanical the oligomer with the lower molecular weight. The oligo-
properties are improved [12]. Unfortunately, it was not ester diols produced were mixed with maleic anhydride,
possible to measure the mechanical properties of the phthalic anhydride and propylene glycol to form unsat-
UPR-3 resin since its high viscosity did not allowed urated polyester resins. These were subsequently diluted
the production of thin lms required for the analysis. with styrene and cured at ambient temperature using the
benzoyl peroxide/N,N-dimethyl-p-toluidine system. The
curing kinetics of this system was investigated and
4. Conclusion the results obtained from the glycolyzed products were
compared to corresponding of a typical alkyd resin rec-
In this research, the glycolysis of recycled PET for the ipe. The replacement of DEG with the glycolyzed PET
production of alkyd resins used as coatings was investi- products gave higher curing rates completed in shorter
210 G.P. Karayannidis et al. / European Polymer Journal 41 (2005) 201210

time periods. However, the elongation at the break point [13] Farahat MS, Abdel-Azim A-AA, Abdel-Raowf ME.
of the resin was higher and the tensile strength lower. Macromol Mater Eng 2000;283:1.
[14] Farahat MS. Polym Int 2002;51:183.
[15] Aslan S, Immirzi B, Laurienzo P, Malinconico M,
References Martuscelli E, Volpe MG, et al. J Mater Sci 1997;32:
2329.
[1] PETCORE Report 2002, 4(4). Available in www. [16] Radenkov Ph, Radenkov M, Grancharov G, Troev K. Eur
Petcore.com. Polym J 2003;39:1223.
[2] Scheirs J. Polymer recycling. Sussex: J. Wiley & Sons; [17] Chen C-H, Chen C-Y, Lo U-W, Mao C-F, Liao W-T
1998. J. Appl. Polym. Sci. 2001;80:943.
[3] Achilias DS, Karayannidis GP. Proceedings of the inter- [18] Mansour SH, Ikladius NE. Polym Test 2002;21:497.
national conference Protection and restoration of the [19] Pimpan V, Sirisook R, Chuayjuljit S. J Appl Polym Sci
environment VI, Skiathos, 15 July 2002. p. 92536. 2003;88:788.
[4] Kosmidis VA, Achilias DS, Karayannidis GP. Macromol [20] Kawamura C, Ito K, Nishida R, Yoshihara I, Numa N.
Mater Eng 2001;286:640. Prog Org Coat 2002;45:185.
[5] Karayannidis GP, Chatziavgoustis A, Achilias DS. Adv [21] Duliban J. Macromol Mater Eng 2001;286:624.
Polym Technol 2002;21:250. [22] Lucas JC, Borrajo J, Williams RJJ. Polymer 1993;34:3216.
[6] Vaidya UR, Nadkarni VM. Ind Eng Chem Res 1987;26: [23] Lee DS, Lee SH, Kim JH. Polym Int 1997;44:143.
194. [24] Eisenberg P, Lucas JC, Williams RJJ. J Appl Polym Sci
[7] Vaidya UR, Nadkarni VM. J Appl Polym Sci 1989;38: 1997;65:755.
1179. [25] Delahaye N, Marais S, Saiter JM, Metayer M. J Appl
[8] Baliga S, Wong WT. J Polym Sci Part A: Polym Chem Polym Sci 1998;67:695.
1989;27:2071. [26] Vilas J, Laza JM, Garay MT, Rodriguez M, Leon LM
[9] Halacheva N, Novakov P. Polymer 1995;36:867. J. Appl. Polym. Sci. 2001;79:447.
[10] Suh DJ, Park OO, Yoon KH. Polymer 2000;41:461. [27] Lu M, Kim S. J Appl Polym Sci 2001;80:1052.
[11] Farahat MS, Nikles DE. Macromol Mater Eng 2001;286: [28] Simitzis J, Zoumpoulakis L, Soulis S. Polym Int
695. 2002;51:308.
[12] Farahat MS, Nikles DE. Macromol Mater Eng 2002;287: [29] Achilias DS, Sideridou ID. J Macromol Sci-Part A: Pure
353. Appl Chem 2002;A39:1435.

View publication stats