High Performance Polymers

http://hip.sagepub.com/

Preparation of poly(butylene terephthalate)/modified organoclay nanocomposite via in-situ

polymerization: Characterization, thermal properties and flame retardancy
Nina Heidarzadeh, Mehdi Rafizadeh, Faramarz Afshar Taromi and Hossein Bouhendi
High Performance Polymers 2012 24: 589 originally published online 5 July 2012
DOI: 10.1177/0954008312446646

The online version of this article can be found at:

http://hip.sagepub.com/content/24/7/589

Published by:

http://www.sagepublications.com

Additional services and information for High Performance Polymers can be found at:

Email Alerts: http://hip.sagepub.com/cgi/alerts

Subscriptions: http://hip.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://hip.sagepub.com/content/24/7/589.refs.html

>> Version of Record - Nov 1, 2012

OnlineFirst Version of Record - Jul 5, 2012

What is This?

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
Article

High Performance Polymers

24(7) 589–602
Preparation of poly(butylene ª The Author(s) 2012
Reprints and permission:

terephthalate)/modified organoclay sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/0954008312446646
hip.sagepub.com
nanocomposite via in-situ
polymerization: Characterization,
thermal properties and flame
retardancy

Nina Heidarzadeh1, Mehdi Rafizadeh1,

Faramarz Afshar Taromi1 and Hossein Bouhendi2

Abstract
Nanocomposites of poly(butylene terephthalate) (PBT)/modified montmorillonite organoclay were prepared via in-situ
polymerization of terephthalic acid and butanediol in the presence of different clay content. Cloisite 30B, a commercially
available amino modified montmorillonite, was modified with 3-aminopropytriethoxysilane (APS) through a silylation
reaction. Morphology and properties of all samples were investigated using X-ray diffraction (XRD), transmission electron
microscopy (TEM), Fourier-transform infrared (FTIR) spectroscopy, proton nuclear magnetic resonance (1HNMR),
differential scanning calorimetry (DSC), and thermogravimetric analysis (TGA). The extent of clay layers in samples was
confirmed by XRD. Results of XRD and TEM showed that an exfoliation structure was produced in nanocomposites.
Studies of crystallization showed that the presence of nanoclay leads to an increase in the crystallization rate and enhances
the thermal stability of nanocomposites. Crystallization kinetics were described by the Avrami equation. Crystal growth
was spherulitic. According to dynamic mechanical analysis (DMA), storage modulus of nanocomposites was remarkably
improved compared with homo PBT. Moreover, a shrinkage test was carried out. Results showed a reduction in shrin-
kages along and across the flow direction that means a decrease in free volume. Flammability based on the UL-94 test was
applied to study the flame retardancy effect of nanoclay on the samples.

Keywords
poly(butylene terephthalate) (PBT), nanocomposite, organoclay, thermal properties, in-situ polymerization

1. Introduction triggers clay exfoliation and nanocomposite formation.

Compared with the melt method which faces interface
During the last decade, nanocomposites have attracted a lot
problems in the preparation of polymer nanocomposites,6
of interest regarding their improved properties, such as
in-situ polymerization is a highly attractive technique due
electrical, magnetic, and mechanical properties, thermal
stability and flame retardancy in comparison with homo
polymers and conventional micro-particle composites.1–3 1
Department of Polymer Engineering and Color Technology, Amirkabir
Melt mixing, solution intercalation, sol-gel, and in-situ University of Technology, Tehran, Iran
polymerization are various methods of nanocomposite pre- 2
Resin and Adhesive Group, Iran Polymer and Petrochemical Institute,
paration. Both uniform distribution and dispersion of Tehran, Iran
nanoscale particles in the polymer matrix are important.
Corresponding author:
Furthermore, good interfacial interaction is the key point
Mehdi Rafizadeh, Department of Polymer Engineering and Color
to reach superior properties.4,5 The in-situ polymerization Technology, Amirkabir University of Technology, P.O. Box 15875-4413,
method relies on swelling organoclay with monomer, fol- Tehran, Iran.
lowed by reaction. Chain growth inside clay galleries then Email: [email protected]

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
590 High Performance Polymers 24(7)

to its versatility and compatibility with various reactive

monomers, which could be a valuable process for mass pro-
duction. To improve the compatibility between organophi-
lic polymers and inorganic clay, interlayer cations, such as
Naþ, Caþ, and Kþ, are replaced with alkyl ammonium
cations and enhanced the dispersion.
The low thermal stability of conventional ammonium
ion modified clays is a major drawback for the preparation
of nanocomposites via in-situ polymerization. If the pro-
cessing temperature is high, modified organoclay will
decompose. The polyesters synthesis temperature is nor-
mally higher than 250 C; therefore, the thermal stability
of these kinds of organoclays is a problem. A great deal
of attention has been directed toward the preparation of
Figure 1. Schematic of the polymerization set-up.
high-temperature stable organoclays.7–10 Alkyl ammonium
surfactants are known to degrade at temperatures between
tetrabutyltitanate (TBT) as catalyst via direct esterification
180 and 200 C. Moreover, thermal degradation of
of terephthalic acid (TPA) and 1,4-butanediol (BDO). Dis-
ammonium also leads to discoloration, molecular weight
persion of nano-particles and thermal properties of samples
reduction, and deterioration of the mechanical properties
were investigated by X-ray diffraction (XRD), transmis-
of nanocomposites.11 To improve the thermal stability of
sion electrical microscopy (TEM), DSC, TGA, and shrink-
organoclay, modification by imidazolium,12–14 phospho-
age test. Thermal behavior was studied using DSC in three
nium,15, 16 stibuniuom,17, 18 and silanes19 have been
various cooling and heating rates. Shrinkage test was per-
reported. Ammonium-modified montmorillonite is the
formed to evaluate dimension stability of nanocomposites.
most studied nanoclay.5 Among modifiers, ammonium
UL-94 test carried out to study flame retardancy.
compounds induce better compatibility between the
polymer matrix and silane compound that could provide
better dispersion. However, both modifications could 2. Experimental details
have a synergistic effect.19 Recently, some results have
been published on the use of nano-additives to prepare 2.1. Materials
materials with enhanced mechanical properties, dimen- TPA was supplied by Shahid Toundgoyan Petrochemical
sional stability, and reduced flammability. The improve- Complex (Mahshahr, Iran). BDO was purchased from
ment in properties of nanocomposites has been studied AppliChem Co. TBT, as a polycondensation catalyst, APS,
intensively.10, 11, 20–27 o-dichlorobenzene (as solvent for intrinsic viscosity mea-
Poly (butylene terephthalate) (PBT) is a semicrystalline surement), and ethanol were prepared from Merck Co.
thermoplastic polyester and has many applications such as (Darmstadt, Germany). Cloisite 30B was purchased from
in electronic parts. However, disadvantages such as low southern clay products. Pure phenol as solvent for intrinsic
impact strength and heat distortion temperature limit the viscosity measurement and triflouroacetic acid (TFA) as
applications of PBT. Furthermore, for in-situ synthesis of proton nuclear magnetic resonance (1HNMR) solvent were
PBT/organoclay nanocomposites, a high temperature, purchased from Sharlau Co. (Spain). All of the materials
about 250–260 C, is required. Therefore, thermally stable were used as received.
organoclay is needed to avoid decomposition of alkylam-
monium cations.28
In this work, commercial bis(2-hydoxyethyl) methyl 2.2. Laboratory set-up
hydrogenated tallow ammonium chloride (MT2EtOH) A homemade reaction unit consisting of a one liter stainless
treated montmorillonite, named Clositie30B, was modified steel reactor, a condenser, and a cooling fan and a vacuum
with 3-aminopropyltriethoxysilane (APS).The structure pump. Figure 1 shows schematic of the set-up. Three
and thermal stability of the modified Clositie30B (30B/ PT100 temperature sensors were mounted inside the reactor,
APS) were studied by means of Fourier-transform infrared jacket, and condenser. The reactor pressure was measured by
(FTIR) spectroscopy and thermogravimetric analysis a pressure transducer. Transducer type is ASHCROFT INC
(TGA) in two thermal modes. The first mode that has been model KM-11 (USA). A two-blade-type mixer was used to
presented by the previous studies (non-isothermal) and the mix highly viscous reactive mixture. A 1500 W electrical
second mode was presented in this work (isothermal) to be heater was installed around the reactor. The sealing of the
able to study behavior during polycondensation. Then, reactor top can tolerate 50 atm. Vacuum was applied by a
PBT/modified Clositie30B nanocomposites with 0.5, 1 and JB-85N- 250 vacuum pump from FastVac Company (USA).
3 wt% of clay were synthesized in the presence of Two chilled traps, in series, were used to prevent the entry of

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
Heidarzadeh et al. 591

Table 1. Different conditions of synthesized samples and their molecular weight.

Sample code BDO/TPA Tes (oC) Pes (bar) Timees (h) Tco (oC) Pco (bar) Timeco (h) [Z] Mna Mnb

Homo PBT-1 1.7 230–235 3–3.5 3 260–263 0.13–0.1 3 0.28 7909 7800
Homo PBT-2 1.7 225–230 2.5–3 3.5 250–255 0.13 3.5 0.39 12,532 12,366
Homo PBT-3 1.7 205–210 3.5 3.5 245–250 0.13–0.14 3 0.55 20,212 –
a
Calculated from intrinsic viscosity.
b
Obtained from 1HNMR.

organic materials into the vacuum pump. The traps were 4 cm-1 and 20 scans/min. A total of 1–2 mg of the sample
cooled in an alcohol (iso-propanol) bath at –30 C. All data was mixed with 100–200 mg of KBr to produce samples for
acquisition and control tasks were performed with home- FTIR.
written Visual C#.net code. I/O card model USB-4711A
from AdvanTech (USA) had responsibility for computer
3.2. 1HNMR
interfacing. An injection system was manufactured to add
1
nanoclay during polycondensation. HNMR spectra were recorded on a Brucker Avance
(Switzerland) 400 spectrometer operated at 400 MHz and
25 C using tetramethylsilane (TMS) as a reference. Polye-
2.3. Modification of Cloisite 30B
sters were dissolved in TFA, and a small amount of CDCl3
Upon modification, Cloisite 30B was dried in a vacuum was added to lock the spectrometer. About 20 or 60 mg of
oven overnight at 60 C. Consequently, about 1 g of clay polyester were dissolved in 0.8 ml of the aforementioned
(Cloisite 30B) was added into a 25 ml APS/ethanol solution for sample preparation.
containing 1 g of APS. The mixture was stirred and
refluxed for 6 h. The product was washed thoroughly by
ethanol/water solution to remove excessive silane coupling 3.3. Intrinsic viscosity measurement
agent, then dried at 60 C for 24 h. The modified organoclay A 1C type Ubbelohde viscometer was used to determine
obtained was designated as 30B/APS. intrinsic viscosity and molecular weight of polyesters.
Phenol and o-dichlorobenzene was used as solvent and
2.4. Synthesis of PBT nanocomposite measurements were performed at 25 + 0.1 C.

Synthesis consists of three steps. In the first step, TPA and

BDO were mixed and transferred to the reactor with 1.7 3.4. XRD
molar ratio of COOH/OH groups. Table 1 presents process XRD profiles were recorded using an X’Pert MPD
conditions. It is observed that homo PBT3 has the highest model (Netherlands) (tube voltage: 40 kV and tube cur-
molecular weight. Hence, all other samples, including nano- rent: 30 mA) with a monochromatic Cobalt radiation
composites, were synthesized under the same conditions as (l¼1.7902 Å). Samples were scanned from 2y¼ 2o up
homo PBT3. The paste was mixed for 30 min at 45 C. Then, to 35o at a scanning rate of 1o/min using a reflection
temperature and pressure were increased to 205 C and mode.
3.5 bar, respectively. Esterification started with a small drop
in temperature. Water was gathered every 15 min as an indi-
cation of extent of reaction. Esterification was completed 3.5. TEM
whenever water production stopped. Then, atmospheric TEM images were taken using EM10C, Carl Zeiss Co.
pressure was applied and mixture of sonicated 30B/APS in (Germany) with a 60 kV acceleration voltage. The sample
BDO and catalyst was injected into the reactor. Temperature was cryomicrotomed into an ultrathin slice of about 100 nm
was increased to 245–250 C and vacuum was applied. This with a diamond knife.
step lasts 3 h. Finally, the pressure was increased to ambient
pressure, and mixture was purged by nitrogen inlet.
3.6. DSC
DSC was used to study thermal and non-isothermal crystal-
3. Measurements and characterization lization behaviors of the nanocomposites. A METTLER
TOLEDO DSC (Switzerland) was used under a nitrogen
3.1. FTIR atmosphere. About 10–11 mg of sample was first heated
FTIR spectra were recorded in the spectral range 4000–500 up to 260 C and held for 5 min to eliminate thermal history;
cm-1 using a FTIR spectrometer BOMEM model (Canada) then it was cooled to 25 C and subsequently reheated up to
at room temperature, with KBr pellets, resolution (RES) ¼ 260 C at a heating rate of 5, 10, and 15 C/min.

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
592 High Performance Polymers 24(7)

3.7. TGA
TGA was performed under a nitrogen atmosphere using a
Pyris 1 TGA Thermogravimetric Analyzer (UK) to investi-
gate the thermal stability of the organoclays, homo PBT,
and nanocomposites. The samples were heated at a rate
of 10oC/min.

3.8. Dynamic mechanical analysis

Thermo-mechanical properties of PBT nanocomposites
were investigated by means of a TA Instrument DMA983
model (USA) at 1 Hz and a scanning rate of 3 K/min
between 100 C and 180 C. Sample size was 45  13  3
mm3 in length, width, and thickness, respectively.

3.9. Shrinkage Figure 2. FTIR spectra of (a) Cloisite 30B and (b) 30B/APS.
The effect of nanoclay on shrinkage behavior of the polye-
sters was studied according to ASTM D955 test. Part
shrinkage across and along the flow direction were calcu-
lated based on the following equations:
100
Sw ¼ ðWm  Ws Þ  ð1Þ
Wm
where Sw , Wm and Ws are shrinkage percentage, mold, and
sample dimension across the flow direction, respectively,
and
100
Sl ¼ ðLm  Ls Þ  ð2Þ
Lm
where Sl , Lm and Ls are shrinkage percentage, mold, and
sample dimension along the flow direction, respectively.
Figure 3. Reaction schematic of (a) Cloisite 30B and (b) APS.
Shrinkage in the width of the sample is presented below:
100 2853 cm-1 determine –CH stretching bond and the peak
St ¼ ðTm  Ts Þ  ð3Þ
Tm at 1470 cm-1 corresponds to –CH2 bending in Cloisite
where St , Tm and Ts are shrinkage percentage, mold, and 30B. Therefore, there is intercalation of alkyl ammonium
sample dimension in the width, respectively. ions among the silicate layers. The peak at 1643 cm-1 rep-
resents OH bending vibrations of the adsorbed water. The
broad peak over the range from 1200 to 1000 cm-1 with the
3.10. Flame retardancy maximum at 1037 cm-1 is due to the Si–O in plane stretch-
The UL-94 horizontal test was carried out based on the ing. The relative peak intensity at around 3400 and 1037
standard ASTM D635 in order to study flammability and cm-1 of Cloisite 30B is much larger than of 30B/APS.
flame retardancy. The test was carried out on a 3 mm thick Therefore, 30B/APS has less hydroxyl groups than Cloisite
sample. 30B due to the reactions between organosilane and hydro-
xyl groups in the hydrogenated tallow ammonium ions.
Grafting of APS on Cloisite 30B was formed by reactions
4. Results and discussions between the organosilane and Si–OH groups on the silicate
edges; see Figure 3. These results are comparable with
4.1. Characterization and thermal properties of 30B/
reported results.11
APS Figure 4 represents XRD patterns of Cloisite 30B and
FTIR spectra of Cloisite 30B and 30B/APS are shown in 30B/APS. XRD pattern of Cloisite 30B shows a peak at
Figure 2. The peak at 3634 cm-1 corresponds to the hydro- 2y ¼ 5.87o, which corresponds to an interlayer distance
xyl stretching that is bonded to the aluminum and magne- of 17.48 Å. However, the pattern of 30B/APS gives a sig-
sium in Cloisite 30B. Also strong peaks at 2928 cm-1 and nificant peak at 2y ¼ 2.56o. The increase of interlayer

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
Heidarzadeh et al. 593

Figure 4. XRD patterns of Cloisite 30B and 30B/APS.

Figure 6. Isothermal (a) TGA and (b) DTG of 30B/APS.

thermal decomposition occurs at a temperature above

250 C. There are shifts of exothermic peaks in DTG from
273 and 365 C for Cloisite 30B to 299 and 402 C for 30B/
APS, respectively. Therefore, there is an enhancement in
thermal stability.
An isothermal TGA test was performed for 30B/APS to
study the thermal stability during the polycondensation
step. Figure 6(a) and (b) shows the isothermal TGA and
DTG of the 30B/APS. In the first 20 min, less than 5% of
weight is lost. Hence, there is enough time to reach an exfo-
liation structure during the polymerization.

4.2. Characterization of homo PBT

Table 1 represents synthesis conditions of homo PBTs and
their molecular weights. Figure 7 shows the FTIR spectra
Figure 5. Non-isothermal (a) TGA and (b) DTG curves of of homo PBT. The peaks around 2961, 1711 and 728 cm-1
Cloisite 30B and Cloisite 30B/APS. belong to aliphatic C–H, C¼O stretching vibration of the
ester carbonyl group, bending vibration of ¼CH of the
distance of 30B/APS indicates that APS was intercalated phenyl group, respectively.29 The peak at 1277 cm1 is
into clay layers. assigned to the stretching vibration of C–O in the carboxyl
Figure 5(a) and (b) represent non-isothermal TGA and group.30
DTG curves of Cloisite30B and 30B/APS. About 6% The 1HNMR spectrum of homo PBT shows three shifts,
weight loss was observed for 30B/APS in Figure 5(a), which represent three kinds of protons, see Figure 8. The
which is due to decomposition of ammonium ions and peaks at 2.2 and 4.5 ppm correspond to the protons of the
organosilanes.11 DTG spectra, Figure 5(b), show that middle and end methylene group of BDO, respectively.

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
594 High Performance Polymers 24(7)

Figure 9. XRD patterns of homo PBT and PBT nanocomposites,

Figure 7. FTIR spectra of homo PBT. in the range of 2y ¼ 2–8o.

Figure 8. 1HNMR spectra of homo PBT.

Figure 10. XRD patterns of homo PBT and PBT nanocompo-
sites, in the range of the 2y ¼ 5–35o.
The peak at 8.1 ppm corresponds to the proton of the phe-
nyl group. Furthermore, ratios of the three peak areas are
2:2:1. Since the area of each peak is proportional to the 4.3. Morphology and thermal properties
number of proton of the related group, the sample could XRD results for all samples are shown in Figures 9 and 10
be recognized as PBT. The number of repeating units (n) for small and wide angles, respectively. In Figure 9, there is
and molecular weight were calculated based on the follow- a small peak at 10.20o for homo PBT that could be attrib-
ing equations:31 uted to the crystalline structure of the polymer.33 For all
n ¼ IHa =IHd ¼ I8:1ppm =I4:8ppm ð4Þ nanocomposites, there is no recognizable peak in the range
of the 2y ¼ 2–8o that was observed in Figure 4 for
30B-APS. Hence, it was concluded that there is a good dis-
Mn ¼ n  M0 ð5Þ persion of the organoclay layer structure. Wan et al.11
where IHa is total area of the peak of phenyl group protons, observed a peak for PBT nanocomposite with 3% nanoclay
and IHd is total area of the chain end methylene proton. In prepared via melt mixing. Hence, in-situ polymerization
Equation (5), Mn is the number average molecular weight gives better dispersion. XRD patterns of the samples in the
and M0 is the molecular weight of the repeating unit. It is higher diffraction angles (2y ¼ 5–35o) are given in
observed that the measured molecular weight based on Figure 10 in order to study the crystalline structure. For
1 homo PBT, strong diffraction peaks observed at the scatter-
HNMR is close enough to molecular weight calculated
using intrinsic viscosity and Mark–Houwink’s equation:32 ing angles 2y of 16.0, 17.2, 20.6, 23.3, 25.2, 29.3, and 31.1,
which correspond to the reflections from the (011), (010),
½Z ¼ K Mna ¼ 4:37  104 Mn0:72 ð6Þ (111), (100), (111), (101), and (111) planes, respectively.

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
Heidarzadeh et al. 595

all cooling rates. Increase of Tc suggests that interaction

between 30B/APS and matrix occurs.36 It could be attrib-
uted to strong interactions among PBT chains and hydroxyl
groups in organoclay.11 It is observed that Tc for PBT3-
30B/APS has the smallest variation with cooling rate
increase.
Crystallization peak width, DTc, is related to overall
crystallization rate. The heat of fusion, DHc, represents the
extent of crystallinity; therefore, the term DHc/time is the
index of crystallization rate. As the clay content increases,
the crystallization rate increases. The crystallization rate of
PBT3-30B/APS is greater than that of homo PBT for all
cooling rates. There is an increase in DTc with increasing
cooling rate. However, there is a decrease in DTc with
increasing clay content.
These results imply that crystallization of homo PBT is
accelerated by incorporating 0.5–3 wt% 30B/APS. Another
important term is the super-cooling degree, defined as DT
¼ Tm – Tc. Here DT conceptually shows overall crystalliza-
Figure 11. TEM micrographs of PBT1-30B/APS.
tion rate. Hence, less super-cooling degree means higher
overall crystallization rate. It is seen in Table 2 that DT
decreases as clay content increases that could be attributed
Therefore, the crystallites are in the a form.34 Characteris-
to nucleation mechanism.36 The minimum for DT is found
tic peaks of homo PBT were also observed in PBT nano-
with the 5oC/min cooling rate, showing the highest nuclea-
composite and the positions of their peaks remained
tion activity in this rate.
almost unchanged in the presence of 30B/APS. Hence, it
In heating runs, Figure 13(a)–(d), the melting tempera-
is concluded that 30B/APS has no effect on the crystal
ture (Tm) of nanocomposites is less than Tm of homo PBT.
structure of the polymer. This result demonstrates that
The lower Tm was attributed to melting of imperfect crys-
incorporation of 30B/APS into PBT matrix does not change
tals,36 formed as a result of the nucleation effect of 30B/
the crystal structure of PBT nanocomposites. However,
APS during a cooling run or the formation of smaller crys-
homo PBT shows sharper peaks that could be related to the
tals within the amorphous region and rearrangement of dis-
larger size of crystallites.35
ordered or uncompleted crystals by the chains which gain
Further investigation of the nanoclay dispersion could
mobility at temperatures before melting.37 However, there
be provided by TEM. Figure 11 illustrates a prototype of
is no meaningful difference in Tm of nanocomposites. The
a TEM photograph of PBT1-30B/APS. The light gray areas
melting peak width, DTm, decreases with clay content. It is
represent intersections of polymer matrix, and the darker
concluded that the distribution of crystallites in nanocom-
regions are regarded as clay. Therefore, an exfoliated-
posites was narrower than that of homo PBT. Moreover,
intercalated morphology was reached.
there is an increase in DTm with heating rate increase, indi-
The DSC heating and cooling scans of homo PBT and
cating a wider crystal size distribution. Decreasing overall
nanocomposites are shown in Figures 12(a)–(d) and
crystallization rate, constant degree of crystallization,
13(a)–(d), respectively. Related data at different cooling
increasing the number of nuclei with clay content suggest
and heating rates are reported in Table 2. Degree of crystal-
that heterogeneous nucleation happens in nanocomposites
linity (wc) was calculated by the following equation:
that leads to an increase in the number of crystals but in
DHm smaller size. In comparison with a melt mixing process,11
wc ¼ ð Þ  100 ð7Þ nanocomposite with 3 wt% organoclay has smaller DTm
DHm0 ð1  fÞ
reflecting narrower crystal size distribution.
where DHm is measured heat of fusion, DHm0 is the heat of Non-isothermal crystallization kinetics can be directly
fusion of 100% crystalline polymer which is 145 J/g,10 and analyzed by the Avrami equation. Relative crystallinity
f is clay content in polymer. It is observed that degree of as a function of temperature, XT, is defined as
crystallinity does not change significantly by clay content.
TRðtÞ
In Figure 12(a)–(d), it is seen that Tc of all samples shifted ðdHc =dT ÞdT
to lower temperatures with increasing cooling rate, indicat- T ð0Þ
ing a decline in nucleation activity. At a given cooling rate, XT ¼ ð8Þ
R
T1
Tc values of PBT nanocomposites are higher than that of ðdHc =dT ÞdT
homo PBT. PBT3-30B/APS has the maximum Tc value for T ð0Þ

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
596 High Performance Polymers 24(7)

Table 2. DSC test data.

DHc J
Samples Cooling rate (oC min-1) Tc ( C) DTc ( C) time ð g:s Þ DT ( C) wc ð%Þ Tm ( C) DHm ð gJ Þ DTm ( C)

Homo PBT-3 5 188.39 9.41 0.4 28.32 32.3 216.71 44.72 7.7
10 184.49 10.27 0.685 32.51 30.8 217 46.58 8.75
15 183.55 15.78 0.697 34.17 31.2 217.72 49.1 12.24
PBT0.5-30B/APS 5 189.74 8.72 0.46 24.36 32.6 214.1 40.22 4.55
10 184.49 9.51 0.8028 30.34 32 214.83 44.96 7.57
15 187.13 12.36 0.957 27.16 28.7 214.29 41.7 8.38
PBT1-30B/APS 5 193.12 6.61 0.54 19.47 27.7 212.59 47.34 3.88
10 186.49 8.89 0.8595 27.34 30.5 213.83 46.41 5.39
15 186.32 11.08 1.021 26.13 33.4 212.45 48.38 7.16
PBT3-30B/APS 5 199.33 6.22 0.62 13.21 32.1 212.54 46.52 3.84
10 195.83 7.53 1.009 17.83 33 213.66 46.41 5
15 195.4 8.52 1.378 15.71 33.8 211.11 45.28 6.88

Figure 12. DSC curves for different cooling rates of (a) homo PBT, (b) PBT0.5-30B/APS, (c) PBT1-30B/APS and (d) PBT3-30B/APS.

Rt
where T(0) and T(t) are the initial and final temperature ðdHc =dtÞdt
0
of crystallization and dHc =dT is heat flow. Relative Xt ¼ ð9Þ
R
t1
crystallinity could be described as a function of time ðdHc =dtÞdt
(Xt): 0

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
Heidarzadeh et al. 597

Figure 13. DSC curves for different heating rates of (a) homo PBT, (b) PBT0.5-30B/APS, (c) PBT1-30B/APS and (d) PBT3-30B/
APS.

Figure 14 illustrates relative crystallinity (Xt) of samples has the highest crystallization rate at all cooling rates. In
versus time. All curves show a sigmoid trend and the rate general, small deviation of curves, from a straight line,
of crystallization increases with clay content, as seen in indicates the presence of a secondary crystallization
Table 2. To calculate Avrami parameters (n and Zt), a process.39
double-logarithmic form was used: TGA results of samples are presented in Figure 16
and Table 4. As clay content increases initial decompo-
ln½ lnð1  Xt Þ ¼ ln Zt þ n ln t ð10Þ sition temperature, TDi, shifts to higher temperatures.
where Zt is a crystallization rate constant involving both The TDi of PBT3-30B/APS is about 28 C more than the
nucleation and growth rate parameters and the Avrami TDi of homo PBT. According to Table 4, residual
exponent, n, is a constant depending on the type of weight at 600 C increases about 4.5%. It could be
nucleation and growth processes. Figure 15 shows attributed to a barrier role of nanoclay in the diffusion
ln½ lnð1  Xt Þversus ln t. Table 3 gives values for n of volatile products during decomposition. Therefore,
and Zt. For homo PBT, n values are in the range of nanocomposites exhibit higher thermal stability com-
3.8–4.3. In general, n decreases with increasing cooling pared with homo PBT. The following equation was used
rate for all samples. Hence, it is concluded that crystal to quantify thermal stability in certain decomposition
growth is spherulitic, and the nucleation mechanism is percent:
homogeneous.38 By adding 30B/APS, n is around 3.0 for Thermal stability change in x% ¼
all samples, indicating a three-dimensional spherulite crys-
Tdeg ðPBT3  30B=APSÞ  Tdeg ðhomo PBTÞ ð11Þ
tallization mechanism and heterogeneous nucleation.38  100
Here Zt increases with clay content, and PBT3-30B/APS Tdeg ðhomo PBTÞ

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
598 High Performance Polymers 24(7)

Figure 14. Degree of crystallinity (Xc) versus time for (a) homo PBT, (b) PBT0.5-30B/APS, (c) PBT1-30B/APS and (d) PBT3-30B/
APS.

Table 5 gives results for homo PBT and PBT3-30B/APS. direction for PBT is about 0.6%, which is attributed to the
There is a mild decrease in thermal stability change for PBT semicrystalline structure and orientation of polymer
PBT3-30B/APS. chains. However, both shrinkage and difference between
Dynamic mechanical thermal analysis (DMTA) was along and across values decrease with clay content. It is
applied in order to study the thermo-mechanical (dimen- also observed that the maximum shrinkage value is
sion stability at high temperatures) behavior. Storage mod- measured along the flow thickness. It is reported by
ulus (E’) and loss factor (tand) curves are presented in Shelesh-Nezhad and Taghizadeh40 that the maximum cool-
Figure 17 and Table 6. The storage modulus shows a slight ing rate happens at the biggest surface after removing the
increase at ambient temperature, but at high temperatures, sample from mold. Hence, shrinkage along and across the
storage modulus variation is significant. Maximum of tan flow direction are slowed down. Moreover, samples show
d occurs at the glass transition temperature, Tg. It is 0.48%, 0.38%, and 1.45% shrinkage reduction in length,
observed that Tg shifts to higher temperature. tan d can width, and thickness, respectively. Therefore, shrinkage
be used to determine the damping behavior and perceive the decreases with clay content and the difference between
restriction of molecular motion that is related to changes in along and across values also decreases. Furthermore, it could
Tg. As shown in Figure 17 and Table 6, the Tg of nanocom- be concluded that clay particles have homogenous disper-
posites was slightly increased. This result was attributed to sion that is consistent with the XRD and TEM results.
the presence of 30B/APS that greatly reduces the mobility Figure 18 shows shrinkage for homo PBT and PBT3-
of the molecules.39 In Figure 17, it is seen that E’ decreases 30B/APS after 24 h. PBT3-30B/APS presents faster shrink-
with temperature and increases with clay content. Moreover, age in comparison with homo PBT due to heterogeneous
E’ of nanocomposites are higher than that of homo PBT. crystallization. PBT3-30B/APS shows less shrinkage. This
Table 7 shows effect of clay content on shrinkage along result is comparable with DSC results.
and across the flow direction after 24 h. It is observed that For further investigation, flame retardancy of samples was
shrinkage difference between along and across the flow measured based on the UL-94 horizontal test. Figure 19

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
Heidarzadeh et al. 599

Figure 15. ln[ln(1X(t))] versus lnt for (a) homo PBT, (b) PBT0.5-30B/APS, (c) PBT1-30B/APS and (d) PBT3-30B/APS.

Table 3. The Avrami parameters for polymers.

Cooling rate
(oC min-1) n Zt

Homo PBT 5 4.3 0.002

10 4 0.004
15 3.8 0.013
PBT0.5-30B/APS 5 3.8 0.023
10 3.4 0.03
15 3.1 0.037
PBT1-30B/APS 5 3.6 0.045
10 3.4 0.05
15 3 0.076
PBT3-30B/APS 5 3.7 0.085
10 3.6 0.187
15 3.4 0.23
Figure 16. TGA curves of homo PBT and PBT nanocomposites.

shows ignition time versus clay content. Table 8 gives on the surface of the sample act as insulators and
ignition time and rate. It is observed that not only igni- decrease the mass loss rate.
tion time but also ignition rate were decreased. Owing to
having high aspect ratio and polar groups, Cloisite 30B
is miscible with polyesters and could provide the best 5. Conclusions
dispersion in comparison with other mineral clays. Cloisite 30B was modified by APS through a silylation
Therefore, nanoparticles could prevent flame from reaction. APS was an effective factor in enhancing the
spreading over the sample because carbon–silicate layers Cloisite 30B thermal stability. Homo PBT and its

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
600 High Performance Polymers 24(7)

Table 4. TGA data of nanocomposites. Table 7. Shrinkage test after 24 h.

Clay content (%) TiD (%)WtR 600 C o

Shrinkage Shrinkage Shrinkage
along the across the along the flow
0 305.37 3.9225 Samples flow (%) flow (%) thickness (%)
0.5 315.37 6.3217
1 318.46 8.3826 Homo PBT-3 2.45 1.83 4.89
3 333.91 8.5453 PBT0.5-30B/APS 2.33 1.81 3.83
PBT1-30B/APS 2.16 1.67 3.55
PBT3-30B/APS 1.97 1.45 3.44
Table 5. Thermal stability change for PBT3-30B/APS.

Degradation Degradation Thermal

temperature temperature for stability
Decomposition for homo PBT PBT3-30B/APS change
(%) (oC) (oC) (oC)

5 354.12 362.7 2.42

10 371.21 380.06 2.38
20 399.88 408.61 2.18
40 422.27 431.33 2.15
60 434.05 443.22 2.11
80 445.06 454.17 2.05

Figure 18. Shrinkage trends of homo PBT and PBT3-30B/APS

during 24 h.

Figure 17. DMA curves of homo PBT and PBT nanocomposites.

Table 6. Storage modulus and loss factor (tan d).

Storage modulus Storage
at ambient modulus at
temperature 109 170 C 107
Samples Tg (at 25 C) (at 170 C) tan d

Homo PBT-3 67.60 1.32 4.90 0.036 Figure 19. Homo PBT and PBT nanocomposites ignition time
PBT0.5-30B/APS 69.17 1.40 5.43 0.035 trend regarding clay content.
PBT1-30B/APS 71.04 1.48 6.12 0.043
PBT3-30B/APS 71.8 1.56 6.47 0.047
XRD patterns of samples were studied in 2y¼ 2–35o. It was
observed that there is no change in crystal structure of poly-
nanocomposites were synthesized via direct esterification mer in the presence of nanoclay. DSC test showed an
in a homemade laboratory scale reactor. Cloisite 30B increase in rate of crystallization with clay content. The
modified APS was homogenized in a PBT matrix during value of DT decreases with clay content. Therefore, more
polycondensation because of silylation reaction. Non- nucleation was occurred. By adding 30B/APS, the Avrami
isothermal TGA has shown that 30B/APS has better ther- parameter, n, is around 3.0 for all samples. Therefore, the
mal stability. Isothermal TGA proved that 30B/APS has crystallization structure is three-dimensional spherulite
acceptable thermal stability for the polycondensation step. with heterogeneous nucleation. TGA results of samples

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
Heidarzadeh et al. 601

Table 8. UL-94 test data. montmorillonite clays and clay/PBT composites. J Appl
Polym Sci 2008; 110: 550.
Ignition Ignition
12. Awad WH, Gilman JW, Nydena M, et al. Thermal degrada-
samples time (s) rate (mm/min) classification
tion studies of alkyl-imidazolium salts and their application
Homo PBT-3 21.6 25 HB in nanocomposites. Thermochim Acta 2004; 409: 3.
PBT0.5-30B/APS 17.9 19 HB 13. Gilman JW, Awad WH, Davis RD, et al. Polymer/layered
PBT1-30B/APS 15.3 16.2 HB silicate nanocomposites from thermally stable
PBT3-30B/APS 14 13.9 HB trialkylimidazolium-treated montmorillonite. Chem Mater
2002; 14: 3776.
14. Ngo H, Compte KL, Hargens L and McEwen AB. Thermal
show that nanocomposites have better thermal stability.
properties of imidazolium ionic liquids. Thermochim Acta
PBT3-30B/APS has faster shrinkage behavior that is attrib-
2000; 357–358: 97.
uted to heterogeneous nucleation and uniform dispersion. A
15. Chang JH and Kim SJ. Poly(trimethylene terephthalate)
flammability test proved that the nanoclay reduced the igni-
nanocomposite fibers by in situ intercalation polymerization:
tion time. A flame retardancy test showed that the ignition
thermo-mechanical properties and morphology (I). Polymer
time and rate were decreased by up to 3%. Therefore, orga-
2004; 45: 5171.
noclay prevents flame spreading.
16. Xie W, Xie R, Pan W, et al. Thermal stability of quaternary
phosphonium modified montmorillonites. Chem Mater
Funding 2002; 14: 4837.
This research received no specific grant from any funding agency 17. Xie W, Gao Z, Pan W, Vaia R, Hunter D and Singh A.
in the public, commercial, or not-for-profit sectors. Characterization of organically modified montmorillonite
by thermal techniques. Polym Mater Sci Eng 2000; 82: 284.
18. Wang D and Wilkie ChA. A stibonium-modified clay and
References its polystyrene nanocomposite. Polym Degrad Stab 2003;
1. Kojima Y, Usuki A, Kawasumi M, et al. Mechanical proper- 82: 309.
ties of nylon 6-clay hybrid. J Mater Res 1993; 8: 1185. 19. Liu H, Zhang W and Zheng S. Montmorillonite intercalated
2. Lan T and Pinnavaia TJ. Clay-reinforced epoxy nanocompo- by ammonium of octaaminopropyl polyhedral oligomeric
sites. Chem Mater 1994; 6: 2216. silsesquioxane and its nanocomposites with epoxy resin.
3. Yao X, Tian X, Zhang X, et al. Preparation and characteriza- Polymer 2005; 46: 157.
tion of poly(butylene terephthalate)/silica nanocomposites. 20. Archondouli PS and Kalfoglou NK. Compatibilization and
Polym Eng Sci 2009; 49: 799. properties of PBT/PU polymeric alloys. Polymer 2001; 42:
4. Liu WT, Tian XY, Cui P, Li Y, Zheng K and Yang Y. Pre- 3489.
paration and characterization of PET/silica nanocomposites. 21. Zhang AL, Zhang BY and Feng ZL Compatibilization by
J Appl Polym Sci 2004; 91: 1229. main-chain thermotropic liquid crystalline ionomer of blends
5. Qu MH, Wang YZ, Wang C, Ge XG, Wang DY and Zhou Q. of PBT/PP. J Appl Polym Sci 2002; 85: 1110.
A novel method for preparing poly(ethylene terephthalate)/ 22. Wu DF, Zhou CX and Zhang M. Effect of clay on immiscible
BaSO4 nanocomposites. Eur Polym J 2005; 41: 2569. morphology of poly(butylene terephthalate)/polyethylene
6. Formes TD, Hunter DL and Paul DR. Effect of sodium mon- blend nanocomposites. J Appl Polym Sci 2006; 102: 3628.
tmorillonite source on nylon 6/clay nanocomposites. Polymer 23. Chen XQ, Xu JJ, Lu HB and Yang YL. Isothermal crystalli-
2004; 45: 2321. zation kinetics of poly(butylene terephthalate)/attapulgite
7. Zhu J, Morgan AB, Lamelas FJ and Wilkie CA. Fire proper- nanocomposites. J Polym Sci Part B: Polym Phys 2006;
ties of polystyreneclay nanocomposites. Chem Mater 2001; 44: 2112.
13: 3774. 24. Chisholm BJ, Moore RB, Barber G, et al. Nanocomposites
8. Zhu J, Uhl FM, Morgan AB and Wilkie CA. Studies on the derived from sulfonated poly(butylene terephthalate). Macro-
mechanism by which the formation of nanocomposites molecules 2002; 35: 5508.
enhances thermal stability. Chem Mater 2001; 13: 4649. 25. Xiao JF, Hu Y, Wang ZH, Tang Y, Chen ZY and Fan WC.
9. Saujanya C, Imai Y and Tateyama H. Structure and thermal Preparation and characterization of poly(butylene terephtha-
properties of compatibilized PET/expandable fluorine mica late) nanocomposites from thermally stable organic-
nanocomposites. Polym Bull 2002; 49: 69. modified montmorillonite. Eur Polym J 2005; 41: 1030.
10. Jin-Hae Changa JH, An YU, Kim SJ and Im S. Poly(butylene 26. Chang J, Mun MK and Kim J. Synthesis and characterization
terephthalate)/organoclay nanocomposites prepared by in of poly(butylene terephthalate)/mica nanocomposite fibers
situ interlayer polymerization and its fiber (II). Polymer via in situ interlayer polymerization. J Appl Polym Sci
2003; 44: 5655. 2007; 106: 1248.
11. Wan Ch, Bao X, Zhao F, Kandasubramanian BP and 27. Li XC, Kang TY, Cho WJ, Lee JK and Ha CS. Preparation
Duggan M. Morphology and properties of silane-modified and characterization of poly(butyleneterephthalate)/

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014
602 High Performance Polymers 24(7)

organoclay nanocomposites. Macromol Rapid Commun II. Nonisothermal crystallization. J Polym Sci Part B: Polym
2001; 22: 1306. Phys 2008; 46: 564.
28. Xie W, Gao ZM, Pan WP, Hunter D, Singh A and Vaia R. 35. Das K, Ray D, Banerjee I, et al. Crystalline morphology of
Thermal degradation chemistry of alkyl quaternary ammo- PLA/clay nanocomposite films and its correlation with other
nium montmorillonite. Chem Mater 2001; 13: 2979. properties. J Appl Polym Sci 2010; 118: 143.
29. Liu L, Qi Z and Zhu X. Studies on nylon 6/clay nanocompo- 36. Broza G, Kwiatkowska M, Roslaniec Z and Schulte K. Pro-
sites by melt-intercalation process. J Appl Polym Sci 1999; cessing and assessment of poly(butylene terephthalate) nano-
71: 1133. composites reinforced with oxidized single wall carbon
30. ShuYong J, YuRong R, Dan Zh, Jing H, Yi Z and GongYing nanotubes. Polymer 2005; 46: 5860.
W. Stannous oxalate: An efficient catalyst for poly(trimethy- 37. Oburoglu N, Ercan N, Durmus A and Kasgoz A. Effects of
lene terephthalate) synthesis. Sci China Ser B-Chem 2008; filler type on the nonisothermal crystallization kinetics of
51: 257. poly(butylene terephthalate) (PBT) composites. J Appl Polym
31. Saint-Loup R, Jeanmaire TH, Robin J and Boutevin B. Synth- Sci 2012; 123: 77.
esis of (polyethylene terephthalate/polye-caprolactone) copo- 38. Wu D, Zhou Ch, Fan X, Mao D and Bian Zh. Nonisothermal
lyesters. Polymer 2003; 44: 3437. crystallization kinetics of poly(butylene terephthalate)/mon-
32. Vincent L, Connolly SN, Dolan F, Willcocks PH, Jayaweera tmorillonite nanocomposites. J Appl Polym Sci 2006; 99: 3257.
SAA and Pendlebury R. Determination and comparison of the 39. SungYeon Hwang, Khaydarov AA, JunYong Park, EuiSang
plane stress essential work of fracture of three polyesters Yoo and SeungSoon Im. Effects of clay modification on crys-
PET, PPT and PBT. J Therm Anal Calorimetry 2006; 86: 147. tallization behavior, physical, and morphological properties
33. Scheirs J and Long TE. Modern Polyesters: Chemistry and of poly(butylene terephthalate) nanocomposites. Macromol
Technology of Polyesters and Copolyester. New York: John Res 2011; 19: 699.
Wiley & Sons, Ltd, 2003. 40. Shelesh-Nezhad K and Taghizadeh A. Shrinkage behavior
34. Huang JW. Poly(butylene terephthalate)/clay nanocomposite and mechanical performances of injection molded polypropy-
compatibilized with poly(ethylene-co-glycidyl methacrylate). lene/talc composites. Polym Eng Sci 2007; 47: 2124.

Downloaded from hip.sagepub.com at Scientific library of Moscow State University on February 20, 2014