Preparation of multifunctional supported metallocene catalyst using organic multifunctional modifier for synthesizing polyethylene/clay nanocomposites via in situ intercalative polymerization

Article abstract of DOI:10.1016/j.polymer.2010.05.041

A new type of multifunctional ammonium modifier with carbonyl group and vinyl group was synthesized to prepare multifunctional montmorillonites (F-MMTs), which were used as multifunctional catalyst supports for in situ ethylene polymerization. High loading of metallocene catalyst in the galleries of F-MMT had been achieved due to the presence of carbonyl group in the multifunctional modifier. XRD profiles and TEM images showed that polyethylene/montmorillonite (PE/F-MMT) nanocomposites with exfoliated structure could be synthesized using the intercalated catalyst described above, even when the content of MMT was very high (more than 15.1 wt%). The as-produced PE/F-MMTs nanocomposites were composed of flower-like particles with a diameter of about 5 μm. A thermal stable monoclinic phase was observed in PE/F-MMT nanocomposites. Comparatively, the resultant PE/F-MMT nanocomposites showed low gas permeability. Interfacial interaction between PE matrix and F-MMT was enhanced due to the chemical linking between the two components via copolymerization of ethylene with vinyl group of F-MMT. Thus the resultant PE/F-MMT nanocomposites showed good structural stability.

Full text of DOI:10.1016/j.polymer.2010.05.041

Polymer 51 (2010) 3416e3424
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Preparation of multifunctional supported metallocene catalyst using organic
multifunctional modifier for synthesizing polyethylene/clay nanocomposites
via in situ intercalative polymerization
,
,
,
,
Changyi Ren a b, Xiaohua Du a b, Li Ma a, Yanhui Wang a, Jun Zheng a b, Tao Tang a
*
a State Key Laboratory of Polymer Physics and Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, China
b Graduate School of the Chinese Academy of Sciences, Beijing 100039, China
a r t i c l e i n f o
a b s t r a c t
Article history:
A new type of multifunctional ammonium modifier with carbonyl group and vinyl group was synthe-
sized to prepare multifunctional montmorillonites (F-MMTs), which were used as multifunctional
catalyst supports for in situ ethylene polymerization. High loading of metallocene catalyst in the galleries
of F-MMT had been achieved due to the presence of carbonyl group in the multifunctional modifier. XRD
profiles and TEM images showed that polyethylene/montmorillonite (PE/F-MMT) nanocomposites with
exfoliated structure could be synthesized using the intercalated catalyst described above, even when the
content of MMT was very high (more than 15.1 wt%). The as-produced PE/F-MMTs nanocomposites were
Received 22 December 2009
Received in revised form
18 May 2010
Accepted 21 May 2010
Available online 1 June 2010
Keywords:
Catalyst
composed of flower-like particles with a diameter of about 5 mm. A thermal stable monoclinic phase was
observed in PE/F-MMT nanocomposites. Comparatively, the resultant PE/F-MMT nanocomposites
showed low gas permeability. Interfacial interaction between PE matrix and F-MMT was enhanced due to
the chemical linking between the two components via copolymerization of ethylene with vinyl group of
F-MMT. Thus the resultant PE/F-MMT nanocomposites showed good structural stability.
Ó 2010 Elsevier Ltd. All rights reserved.
Clay
Nanocomposite
1. Introduction
interlayer of MMTs, i.e. ion-exchange with cationic metallocene
species [9], direct intercalation of catalyst into modified MMTs [3],
The in situ intercalative polymerization for preparing polyolefin/
montmorillonite (MMT) nanocomposites is carried out in the
presence of an intercalated catalyst within modified MMTs [1e7].
This method has been believed to be more desirable in terms of
effectiveness to form nanostructures than solution or melt polymer
intercalation assisted by functional polyolefin compatibilizer [8]. In
this case, olefin polymerization generally takes place within the
interlayer of MMTs. As polyolefin chains grow, the MMT sheets are
exfoliated gradually. However, two problems are usually encoun-
tered for this method. On one hand, as the driving force for the
exfoliation of MMT during in situ intercalative polymerization of
olefin results from the push force of gradual growing polyolefin
chains within the interlayer of MMTs, an adequate amount of
catalyst must be fixed in the gallery of MMTs. This depends on how
much the catalysts diffuse into the interlayer of MMTs where they
should be anchored. Thus organic modification of MMTs is neces-
sary to provide enough interlayer distance for the diffusion of
catalysts [3]. There are three routes for fixing catalysts within the
and supporting with hydroxyl group exposed on the surface of
modified MMT sheets via MAO pretreatment [4,7,10,11]. There are
plenty of highly reactive structure CH3eAle in MAO, which would
react with reactive groups such as eO*H or eCO*e to form chemical
bond eO*eAle or eC(CH3) eO*eAle; then the catalysts would be
supported onto the support via MAO that had been anchored onto
the catalyst support. The last method is widely applied; however,
most of the hydroxyl groups of MMTs are on the edges, not inside of
MMT sheets, mainly leading to the polymerization out of the
interlayer of MMT sheets. As a result, the formation of the exfoliated
structure is limited, especially when the content of MMTs is high. To
resolve this problem, a simple way is to introduce more anchoring
sites into the interlayer of MMTs for supporting catalyst, such as
hydroxyl group for the supporting of metallocene catalyst [1,5] and
MgCl2 for the supporting of ZieglereNatta catalyst [12]. On the
other hand, theoretical studies have showed that simple dispersion
of MMTs in polyolefin matrix is thermodynamically unstable in
spite of the claimed successes of the in situ polymerization method
due to immiscibility between polyolefin and clay [1,11,13,14]. Thus
polar groups should be included in polyolefin backbones to
improve their compatibility with MMTs and enhance the structural
stability of nanocomposites formed by in situ intercalative
* Corresponding author. Tel.: þ86 431 85262004; fax: þ86 431 85262827.
E-mail address: ttang@ciac.jl.cn (T. Tang).
0032-3861/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2010.05.041

C. Ren et al. / Polymer 51 (2010) 3416e3424
3417
polymerization [10]. Alternatively, strong interaction can also be
gained by inserting organic modifier into polymer chains [15,16], as
organic modifiers for MMTs have strong interaction with MMTs
through electrovalent interaction. In this case, functionalized
polyolefin itself served as an organic modifier [17].
Therefore, to prepare polyolefin/MMT nanocomposites with
exfoliated structure via in situ intercalative polymerization, it is
requisite to increase the loading of catalyst in the interlayer of
MMTs for promoting exfoliation as well as to enhance the inter-
action in the interface between polyolefin and MMTs for improving
6, 7-tetra-hydro-1-indenyl) zirconium dichloride (rac-Et
(IndH4)2ZrCl2, Boulder Scientific Company, marked as Zr-Cat for
short), methylaluminoxane (MAO: 10 wt% in toluene, Ethyl Corp.)
and ethylene (polymerization grade, Liaoyang Chemical Corp.)
were used without further treatment. Undec-10-enylamine was
synthesized through Gabriel synthesis method using 11-bromo-
undecene, yielding 45%.
2.2. Synthesis
the structural stability. Shin et al. used
u-undecylenylalcohol as
The synthesis route for multifunctional modifier was summa-
rized in Fig. 1a. The preparation of supported catalyst and the in situ
polymerization of ethylene were summarized in Fig. 1b.
a polymerizable modifier to modify MMTs in which an intercalated
transition metal catalyst was then prepared to synthesize poly-
ethylene (PE)/MMTs nanocomposites [18]. However, the detailed
information about modified MMT, intercalated catalyst and copo-
lymerization of ethylene with the polymerizable modifier was not
provided. Thus, to date, further exploration for a synthetic method
to satisfy the both requirements at the same time is obviously
necessary.
Herein, to meet all the needs above simultaneously, we report
the design and synthesis of a novel multifunctional MMT (F-MMT)
by ion-exchange using a new multifunctional ammonium modifier
and the preparation of exfoliated PE nanocomposites based on the
F-MMT. This modifier contained carbonyl group and vinyl group
besides ammonium group. The carbonyl groups could anchor
catalysts within the interlayer of F-MMT, which could also be
reduced to hydroxyl group by MAO to increase the structural
stability; by in situ intercalative polymerization, the vinyl group in
the F-MMT could copolymerize with ethylene to strengthen the
interaction between PE chains and F-MMT. The microstructures
and structural stability of the resultant PE nanocomposites were
investigated.
2.2.1. Synthesis of multifunctional modifier (Compound 4 in Fig. 1a)
The precursor Compound 1 was prepared as follow. A solution of
4-methylacetophenone (6.0 g) in dried CCl4 (150 ml) was added
NBS (8.0 g) and AIBN (0.1 g). After the reaction at 65 ꢀC was
complete, the solid succinimide, which floats upon the surface, was
filtered off and the solvent was removed by a rotary evaporator. The
crude product was purified by recrystallization from CH3COOEt/n-
C6H14 (6/3, v/v; yielding 42%).
The precursor Compound 2 was prepared as follow. Into a dry
250 ml two-necked flask was transferred benzene (100 ml), fol-
lowed by Compound 1 (2 g), Et(OH)2 (1 ml) and TsOH (0.02 g). A
water separator was equipped on the flask. The mixture was
refluxed at 80 ꢀC for 1 day, and the solvent was removed by a rotary
evaporator. The crude product was purified by recrystallization
from benzene (yielding 85%).
The precursor Compound 3 was prepared as follow. 0.6 g of
undec-10-enylamine, 100 ml of CH3CN and 2.0 g of powder K2CO3
were transferred into a dry 250 ml two-necked flask. A condenser
was attached and the reaction mixture was placed in an ice bath.
3.8 g of Compound 2 was dissolved in 50 ml of CH3CN and added
into the flask dropwise. After 0.5 h, the temperature was raised up
to 35 ꢀC to react for 1 day. After reaction the mixture was filtered,
and the filtrate was kept. The crude product was got by evaporating
the solvent, and was purified by thin layer chromatography (TLC,
silica gel) using a mixed solvent of Et2O and n-C6H14 (1/4, v/v;
yielding 53%).
The multifunctional modifier Compound 4 was prepared as
follow. A nitrogen-degassed solution of Compound 3 (3 g), TsOH
(0.12 g) and H2O (0.13 g) in CH3COCH3 (50 ml) was kept at room-
temperature for 3 days. The solvent was removed in vacuo and the
residue was partitioned between Et2O and H2O. The organic phase
2. Experimental section
2.1. Materials
Naþ-montmorillonites (MMTs) used in this work, with a cation
exchange capacity (CEC) of 119 meq/100 g, were from Kunimine Co..
Analytical grade CH3CN, CH3COCH3, K2CO3, Et(OH)2, C2H5OH, TsOH,
Na2SO4, n-C6H14, Et2O, CH3COOEt, benzene, azobisisbutyronitrile
(AIBN), n-bromosuccinimide (NBS) and octadecylamine were used
without further purification. Carbon tetrachloride (CCl4) and
toluene was dried before use. 4-Methylacetophenone (95%
Aldrich), 11-bromo-undecene (95%, Aldrich), rac-ethylene bis (4, 5,
Fig. 1. Synthetic route of multifunctional modifier (Compound 4) (a) and supporting process of catalyst (b).

3418
C. Ren et al. / Polymer 51 (2010) 3416e3424
was kept, and the crude product was got by evaporating the
solvent. The multifunctional modifier Compound 4 was purified by
thin layer chromatography (TLC, silica gel) using a mixed solvent of
Et2O and n-C6H14 (1/3, v/v; yielding 90%).
investigate the morphology of PE nanocomposites. Field emission
scanning electron microscope (FESEM) and energy-dispersive X-ray
(EDX) analysis were used to investigate the morphology and the
composition of the samples on a XL 30 ESEM FEG scanning electron
microscope (Micrio Fei Philips, Holand). Rheological properties
were measured on a PHYSICA MCR 300 at 190 ꢀC as a function of
angular frequency (ranging from 0.01 to 100 rad/s). The permeation
experiments of oxygen through neat PE film and the nano-
composite films were measured with a gas transmission rate (G.T.
R.) measurement apparatus (K-315-N-03) at 30 ꢀC. The thickness of
2.2.2. Preparation of the multifunctional MMT
The resultant multifunctional modifier was protonated with
equal molar concentrated HCl in a mixture solution C2H5OH/H2O
(3/7, v/v) at 80 ꢀC via cation exchanging with the interlamellar
cations of the MMT. The multifunctional MMT (F-MMT) was got
after the sample was washed and freeze-dried.
the samples was about 30
mm.
2.2.3. Preparation of the supported catalysts
3. Results and discussion
A certain amount of F-MMT (0.05e0.40 g) reacted with excess
MAO (10 ml) under the protection of Argon for 4 h at 60 ꢀC in
toluene in a dried Schlenk-type filtration reactor equipped with
a magnetic stirrer. Then the sample was washed with toluene to
remove the excess MAO. The resultant sample was named as MAO-
F-MMT. It was followed by injecting excess of Zr-Cat solution
(1 ꢁ 10ꢂ3 M, 10e15 ml) and reacting at 60 ꢀC for another 4 h. Then
the sample was washed with toluene for three times to get the
supported catalyst, named as Zr-F-MMT.
Fig. 2 shows the 1H NMR spectra of Compound 4 in CDCl3-d1. The
protons corresponding to the chemical shifts are marked. It is
confirmed that the multifunctional modifier precursor contains one
amine group, two carbonyl groups and one vinyl group in its
molecular structure (Fig. 2). Using protonated multifunctional
modifier, multifunctional MMT (F-MMT) was synthesized via ion-
exchange. TGA results showed that F-MMT contained about 31.7 wt
% the multifunctional modifier (marked as wt%FM).
Fig. 1b shows the preparation process of PE/F-MMT nano-
composites by in situ intercalative polymerization. Here Zr-Cat is
supported onto F-MMT (or O-MMT as a control) under the designed
experimental conditions to prepare the intercalated catalysts, i.e.
Zr-F-MMT (or Zr-O-MMT). According to the measurement of ICP,
the Zr content of Zr-F-MMT was 3.1 wt%, which was obviously
higher than that of Zr-O-MMT (0.4 wt%). Probably the carbonyl
groups within the galleries of F-MMT acted as anchoring sites for
MAO and then Zr-Cat during supporting process, which is the
reason why the loading of catalyst in the Zr-F-MMT is much higher
than that in the Zr-O-MMT. XPS spectra were used to investigate
the formation of the supported catalyst (Fig. 3). The first peak of the
XPS spectrum of Zr-Cat at 181.4 eV was attributed to the Zr 3d5/2
binding energy; the second peak at higher binding energy was
attributed to the Zr 3d3/2 binding energy [19]. Owing to the spin-
orbit splits (3d5/2ed3/2), the XPS spectrum of Zr-F-MMT did not
show a clear peak of Zr 3d spectral line. The binding energy of the Zr
3d spectral lines and the spin-orbit splits (3d5/2e3d3/2) in the XPS
spectra of supported catalyst would be a function of catalyst
preparatory methods. Usually, the spin-orbit (3d5/2e3d3/2) splits
into two sets, but the intensity of 3d5/2 peak in either set is stronger
than that of 3d3/2 peak. So the peak of Zr 3d5/2 would be clear due to
strong intensity of the peak; but the peak of Zr 3d3/2 would be hard
to tell [19]. According to Fig. 3, the peak of Zr 3d5/2 showed a rise of
2.4 eV in the binding energy compared to that of Zr-Cat. This
indicates that the Zr-Cat was supported on F-MMTs. Furthermore,
the supporting process was also confirmed by the XRD data. Fig. 4
shows the XRD profiles of F-MMT, MAO-F-MMT and Zr-F-MMT. The
(001) diffraction peak of MMT in the MAO-F-MMT shifted to lower
2.2.4. Ethylene polymerization
The resultant catalyst was transferred to a polymerization
reactor under protection of Argon. 100 ml of toluene and 3 ml of
MAO were added at 60 ꢀC under mechanical stirring, and the
polymerization started when ethylene was introduced into the
reactor. For different samples, polymerization time changed to get
different MMT loading. The polymerization was terminated using
acidified ethanol solution; the precipitated product was collected
after being washed with ethanol and dried in vacuo at 60 ꢀC for
24 h. For comparison, an octadecylammonium modified MMT (O-
MMT) was prepared and the same catalyst was supported on it to
perform in situ polymerization under the same experimental
condition for F-MMT. The supported catalyst was respectively
named as Zr-O-MMT. A neat PE sample (HOMO) was also prepared
for comparison.
2.3. Characterization
Thermal gravimetric analysis (TGA) was carried out on a Per-
kineElmer TGA-7 Series Thermal Analysis System at a heating rate
of 10 ꢀC/min from 50 ꢀC to 700 ꢀC under nitrogen gas. Room-
temperature and high-temperature 1H NMR and 13C NMR spectra
were recorded on a Bruker AM-400 instrument. Molecular weight
and molecular weight distribution of the products were measured
with gel permeation chromatography (GPC, PL-GPC-220) operated
at 135 ꢀC in 1, 2, 4-trichlorobenzene. Before NMR and GPC
measurements, the MMT sheets in PE nanocomposites were
removed by dipping them into concentrated HF solution for a week,
then washed with ethanol to remove the unreacted modifier and
dried. The Zr content of the supported catalysts was determined by
inductively coupled plasma atomic emission spectroscopy (ICP)
with Plasma-Dec(I) of America Leeman Lab. X-ray photoelectron
spectroscopy (XPS) for chemical analysis on a VGESCA Lab MK II
was used to investigate the binding energies of the atoms of Zr-Cat
and the supported catalyst Zr-F-MMT. Room-temperature and
high-temperature wide angle X-ray diffraction (WAXD) were
carried out using a Rigaku model Dmax 2500 with a Cu Ka radiation,
and the scanning rate was 5ꢀ/min. In the high-temperature WAXD
measurement, the temperature rose from 30 ꢀC to 150 ꢀC, and then
cooled down to 30 ꢀC; the heating/cooling rate was 5 ꢀC/min, and
the data collection time was 6 min per scan. Transmission electron
microscope (TEM, JEOL JEM-2010(HR), at 200 kV) was used to
Fig. 2. 1H NMR spectrums of multifunctional modifier Compound 4 in CDCl3-d1.

C. Ren et al. / Polymer 51 (2010) 3416e3424
3419
PDI
Table 1
Results of ethylene polymerization catalyzed by intercalated catalysts.
Samplea
MMTb
wt%
Activity
Mw
10ꢂ6 g PE atmꢂ1
h
ꢂ1∙(mol Zr)ꢂ1
104c
HOMO
F1
F2
F3
F4
Null
3.7
8.0
15.1
17.7
8.2
2.20
22.7
13.2
13.0
12.0
12.2
20.6
2.7
3.0
2.9
3.2
2.8
2.8
0.83 (278)d
0.80 (268)
0.79 (264)
0.84 (280)
0.71 (31)
O1
a
Other conditions: 0.1 MPa ethylene, 100 ml toluene,
3
ml MAO,
temperature ¼ 60 ꢀC.
b
Measured by TGA analysis.
Measured by GPC, the samples containing MMTs was dipped in HF to remove
c
MMT before examine.
d
The parentheses were the activities of supported catalyst: g PE atmꢂ1 ꢂ1∙(g
h
supported catalyst)ꢂ1
.
Fig. 3. XPS spectra of Zr-F-MMT and Zr-Cat.
was not listed). But there was still a clear peak of O-MMT for O1,
indicating that the O-MMT is not completely exfoliated. This may
be due to the low Zr-Cat loading of Zr-O-MMT. TEM images further
proved that both F1 and F3 were of exfoliated structure (Fig. 5a and
b), but O1 was not in spite of a low loading of MMT compared with
F3 (Fig. 5d). In the high-magnification TEM image of F3, individual
MMT sheets could be observed clearly (Fig. 5c). Comparatively,
without the effective intercalation of enough amount of Zr-Cat into
the interlayer of MMT, such as in the case of O-MMT, the large
aggregates of MMT sheets were observed in the O1 sample. These
results confirm that it is necessary to introduce the high amount of
Zr-Cat into the interlayer of MMT in the preparation of PE/MMT
nanocomposites via in situ intercalative polymerization.
angle as compared to that of F-MMT, indicating that MAO has
intercalated into the gallery of F-MMT; the (001) diffraction peak of
MMT in the Zr-F-MMT shifted to lower angle as compared to that of
MAO-F-MMT, indicating that the Zr-Cat has been supported in the
gallery of MAO-F-MMT.
Table 1 shows the polymerization results. Six samples included
one neat PE (HOMO) from homogeneous catalyst, four PE/F-MMT
nanocomposites (F1eF4) from the intercalated catalyst Zr-F-MMT
and one PE/O-MMT nanocomposite (O1) from the intercalated
catalyst Zr-O-MMT. The results showed that the catalytic activities
of all the supported catalysts were lower than that of homogeneous
catalyst, but the catalytic activity of Zr-F-MMT per gram of sup-
ported catalyst was much higher than that of Zr-O-MMT. Compared
to HOMO sample and O1, the molecular weight of PE in the PE/F-
MMT nanocomposites was low due to copolymerization, which is
Very surprisingly, the nascent morphologies of both F2 and O1
showed different superstructures although both the samples con-
tained the same content of MMT (Fig. 6). The as-produced nano-
composite of F2 was composed of flower-like particles with
a diameter of about 5 mm (Fig. 6a), in which there were many layer
similar to the results of copolymerization of ethylene with
[20,21].
a-olefin
structures (Fig. 6b). In contrast, the nascent morphology of O1 was
of aggregated particles (Fig. 6c), and no flower-like structures were
observed (Fig. 6d). This may result from the dispersion of organic
modified MMT and the “morphology duplicating effect” of sup-
ported catalyst [22]. The organoclay would swell in the organic
media at different scales, i.e., interplatelet swelling at nanoscopic
scale; swelling at the microscopic scale to form spherical gels by
percolation of microgels, based on swollen 3e4 platelet tactoids
[23]. In the case of Zr-F-MMT, more Zr-Cat was supported within
the galleries of MMT. Thus almost each Zr-F-MMT sheet served as
the catalyst supporter, and the resultant PE duplicated the
morphology of Zr-F-MMT sheet to form a layer structure. At the
same time, Zr-F-MMT was swelling at the microscopic scale to form
spherical gels in toluene; so the as-produced PE/F-MMT nano-
composite was composed of spherical flower-like particles.
However, in the case of Zr-O-MMT, the loading of Zr-Cat within the
gallery of MMT is low, so Zr-O-MMT tactoids would serve as the
catalyst supporter. As a result, the resultant PE duplicated the
morphology of Zr-O-MMT tactoids to form solid particles. In this
case, MMT platelets are not completely exfoliated during ethylene
polymerization. The detailed formation process for the formation of
the flower-like particles needs to be further investigated. A similar
superstructure was observed in the F3 (Fig. 7). The as-produced
nanocomposite was also composed of flower-like particles with
Fig. 4 also presents the XRD profiles of F2, F3 and O1 samples.
Very strikingly, at the high content of F-MMTs (sample F3, 15.1 wt
%), the (001) diffraction peak of MMT disappeared, meaning that F-
MMT is completely exfoliated in PE matrix, without saying F2
(8.0 wt%) and F1 (3.7 wt%; XRD profile, similar to that of F2 and F3,
a diameter of about 5
mm, in which there are many layer structures
(Fig. 7b). EDX image showed a homogeneous dispersion of Si
element on the as-produced particles of F3 (Fig. 7d), which further
confirms the homogeneous dispersion of MMT sheets in the F3
sample.
Fig. 4. XRD profiles of organic modified MMT (F-MMT), MAO treated F-MMTs (MAO-F-
MMT), catalyst supported MMT (Zr-F-MMT) and PE/MMT nanocomposite F2 (8.0 wt%
F-MMT), F3 (15.1 wt% F-MMT) and O1 (8.2 wt% O-MMT) gained by in situ
polymerization.

3420
C. Ren et al. / Polymer 51 (2010) 3416e3424
Fig. 5. TEM images of PE nanocomposites. (a) F1; (b) F3; (c) TEM image of F3 at high-magnification, individual F-MMT sheets were marked by arrows; (d) O1.
Fig. 8 shows the WAXD patterns of as-produced O1, F2 and F3.
WAXD pattern of as-produced O1 did not have the (010) monoclinic
reflection. This indicates that ethylene polymerization in the gallery
of F-MMTs should be more drastic than in the gallery of O-MMT,
which is caused by the significant difference in the catalyst loading
of both F-MMTs and O-MMTs. High-temperature WAXD was used
to further investigate this phenomenon (Fig. 9). As mentioned, the
monoclinic phase is metastable, usually it can be eliminated by
thermal treatments. In the high-temperature WAXD measure-
ments, F2 was heated up to 150 ꢀC then cooled down to 30 ꢀC. The
reflection peaks were slightly shifted to lower degree when the
temperature rose, which is due to the expansion of the crystalline
cell. Surprisingly, the peak of (010) monoclinic reflection could be
still observed after the heat treatment. The reason for that is not
clear. It may be due to the high content and good dispersion of F-
MMT platelets, which hinder the crystallites from reorganization
[28].
Besides increasing the catalyst loading, the formation of chem-
ical linking between PE matrix and MMTs in PE/F-MMT nano-
composites was the other important intention using the
multifunctional modifier to modify MMTs. The sample F4 with the
highest loading of MMT (17.7 wt%) was used for NMR character-
ization, considering the insensitive of NMR analysis to the moieties
of low concentration. Fig. 10 shows the 1H NMR (a) and 13C NMR (b)
spectra of F4 (after HF treatment) in 1,1,2,2-tetrachloroethane-d2 at
130 ꢀC. The chemical shifts and their corresponding structures were
The arrows were used to mark the reflections of the nano-
composites. Very interestingly, (110) and (200) orthorhombic
reflections were found in the patterns of F2 and F3, which are
typical reflections of PE matrix, meanwhile (010) monoclinic
reflection was also observed. It is well-known that the stable
orthorhombic phase of PE can be transformed into the metastable
monoclinic phase by stress [24,25]. This process is called
martensitic transformation. In the in situ intercalative polymeri-
zation for preparing PE/MMT nanocomposites, it is believed that
the exfoliation of MMT platelets are driven by the stress, which
should be formed in situ when more and more PE chains are
synthesized in the limited space of intergalleries of MMTs. There
were other reports that monoclinic crystals were formed in the PE
sample synthesized by using heterogeneous catalysts [26,27]. The
presence of monoclinic crystals was believed to be consistent with
a deformation process, arising from impingement of growing
polymer particles [26]. The concentration of monoclinic crystals
could be increased when a-olefin was introduced in the PE back-
bone by copolymerization, as the side chain of the resultant PE
would disturb the crystallization of orthorhombic phase to form
more monoclinic crystals [27]. This may be another reason that
monoclinic phase was found in PE/F-MMT nanocomposites, as the
multifunctional modifiers would be proved to be involved in the PE
backbone by copolymerization in following discussion. But the

C. Ren et al. / Polymer 51 (2010) 3416e3424
3421
Fig. 6. FESEM images of as-produced F2 (a, b) and O1 (c, d) at low-magnification and high-magnification.
marked in the spectra. In Fig. 10a, the chemical shift at
represents the characteristic resonance for the marked proton d. In
Fig. 10b, the chemical shifts at ¼ 131.5 ppm,
¼ 146.7 ppm,
¼ 128.3 ppm and ¼ 118.1 ppm represent the characteristic
d
¼ 7.16 ppm
confirming that carbonyl groups act as the anchoring sites for MAO
and then Zr-Cat during supporting process. Furthermore, by
calculation using the integral intensity of 1H NMR peaks, GPC and
other available data, more detailed information about copolymer-
ization of ethylene with F-MMT could be evaluated, including the
weight percentage of the modifier linked to PE backbone (wt
d
d
d
d
resonance for the marked carbon b, e, d and c, respectively. These
results mean that the modifier molecule has been inserted into PE
backbone. By theoretical calculation and simulation, the change of
chemical shift indicates that the carbonyl groups on the multi-
functional modifier were reduced to hydroxyl group by MAO,
%FMꢂPE), the molar percentage of reacted modifier molecules (RCM),
the number of the modifier molecules (from each MMT sheet)
linked to PE backbone (NFMꢂMMT), the number of the modifier
Fig. 7. FESEM images of as-produced F3 at low-magnification (a) and high-magnification (b) and EDX image of Si element (d) corresponding to image (c).

3422
C. Ren et al. / Polymer 51 (2010) 3416e3424
Fig. 8. WAXD patterns of as-produced O1 (a), F2 (b) and F3 (c); the arrows marked the (110), (200) orthorhombic reflections and (010) monoclinic reflection of PE.
molecules linked to each PE chain (NFMꢂPE), and the number of PE
chains linked to each MMT sheet (NPEꢂMMT). They could be calcu-
lated by Equations (1)e(5):
AMMT is the surface area of MMT, it is believed to be 700e800 m2/g
[29]. NA represents the Avogadro constant. Asheet represents the
surface area of each sheet of MMT. Supposing the shape of MMTs is
round, then Asheet ¼ 2
p
d2/4, and because the aspect of used MMTs is
MFM ꢁ IFM=4
100e200, hence d (diameter of each sheet) could be 100e200 nm.
The average values of AMMT and d are taken, i.e. 750 m2/g and
150 nm respectively. MnPE is the number-average molecular weight
of PE, which could be calculated by Mw and PDI in Table 1. By
calculation, it turns out that wt%FMꢂPE ¼ 1.6 wt%, RCM ¼ 23.5 wt%,
NFMꢂMMT ¼ 2110, NFMꢂPE ¼ 1.8, and NPEꢂMMT ¼ 1172. As it is well-
known that naturally occurring layered silicates (such as MMTs) are
inhomogeneous across layers [30], these results only fit for the case
of F4. Hence such calculations serve only as an approximate guide
to the copolymerization efficiency in the PE/F-MMT
nanocomposites.
wt%FMꢂPE
¼
(1)
(2)
(3)
(4)
(5)
MFM ꢁ IFM=4 þ MCH2 IPE=2
ð1 ꢂ wt%FꢂMMTÞ ꢁ wt%FMꢂPE
wt%FꢂMMT ꢁ wt%FM
RCM
¼
RFM ꢁ NA ꢁ wt%FM=ð1 ꢂ wt%FM
Þ
NFMꢂMMT
¼
MFMAMMT=Asheet
wt%FMꢂPE=MFM
¼
1=MnPE
NFMꢂPE
The chemical linking provides the strong interfacial interaction
and good compatibility between PE matrix and F-MMTs. Extraction
experiment was performed to demonstrate the interfacial interac-
tion between PE matrix and MMTs in the PE/F-MMT nano-
composites. Fig. 11 shows the EDX patterns of F2 and O1 after
extraction using boiling decalin, and the insets are the corre-
sponding FESEM images. According to the EDX patterns, a large
amount of PE remained in the case of F2 after extraction for 48 h; by
contraries, only a trace amount of PE remained in the case of O1.
NFMꢂMMT
NFMꢂPE
NPEꢂMMT
¼
Here MFM represents the molecular weight of the multifunctional
modifier, MCH represents the molecular weight of methylene unit
2
(CH2) in PE backbone. IFM and IPE are the integral intensities at the
chemical shifts corresponding to proton ii and PE backbone on the
1H NMR spectrum, respectively. They are 1 and 1038 according to
Fig. 10. wt%FM represents the weight percentage of the multifunc-
tional modifier in the F-MMT, which is 31.7 wt%. wt%FꢂMMT repre-
sents the MMT content of the sample, here it is 17.7 wt% for F4.
Fig. 10. 1H NMR (a) and 13C NMR (b) spectrum of concentrated HF treated F4 (17.7 wt%
MMT) in 1,1,2,2 -tetrachloroethane-d2 at 130 ꢀC; the chemical shifts and their corre-
sponding structure are marked.
Fig. 9. High-temperature WAXD patterns of F2 (the temperature raised from 30 ꢀC to
150 ꢀC then cooled down to 30 ꢀC, and the heating/cooling rate was 5 ꢀC/min).

C. Ren et al. / Polymer 51 (2010) 3416e3424
3423
Fig. 12. Plots of log G0, G00 and j
h*j vs. log
u
for HOMO, F2 and O1 at 190 ꢀC.
permeability of PE. However, the gas permeability of a polymer/clay
nanocomposite depends on the dispersion degree and the orien-
tation degree of the clay platelets in the polymer matrix. In this
work, there was no orientation of clay in PE/F-MMT and PE/O-MMT
nanocomposites. Therefore the gas permeability only depends on
the dispersion degree of clay platelets. The permeability coeffi-
cients of oxygen for HOMO, O1 (8.2 wt% O-MMT) and F2 (8. F-MMT)
at 30 ꢀC are listed in Table 2. Compared to pristine PE (HOMO), the
gas permeabilities of O1 and F2 were reduced 67% and 85%,
respectively. The difference demonstrated that the exfoliation
degree of clay platelets in the F2 was much higher than that in the
O1. PE/F-MMT nanocomposites with low gas permeability are
potential in the application, such as food and beverage package, etc.
Annealing experiments via compression molding at high-
temperature were used to monitor the structural stability of the
resultant nanocomposites. After annealing at 170 ꢀC for 20 min,
both WAXD profiles and TEM images showed that the exfoliated
structures were still kept in the F1 and F2 comparing with those of
original samples (Fig. 13). This proves that PE/F-MMT nano-
composites have the good structural stability even at high-
temperature. The main reason for the good structural stability
results from the formation of strong interfacial interaction between
PE chains and MMT sheets due to chemical linking of some PE
chains to the modifier molecules in the F-MMTs. In the PE/F-MMT
nanocomposites, the modifier molecules incorporated into PE
chains act as a bridge. They have strong interaction with PE chains
via covalent bonding and with MMT sheets via electrovalent
interaction. Additionally, according to the 1H NMR analysis of
model reaction between MAO and the modifier precursor, the
carbonyl groups could be reduced to hydroxyl group, which could
further enhance the interaction between the modifier molecules
and the surface of MMT sheets. As a result, the structural stability of
PE/F-MMT nanocomposites can be further improved.
Fig. 11. (a) EDX pattern of F2 after extracting pure PE with boiling decalin for 48 h
(corresponding to the area shown in the FESEM image a1), the insets are FESEM
micrographs of F2 after solvent extraction at low-magnification (a1) and high-
magnification (a2); (b) EDX pattern of O1 after extracting pure PE with boiling decalin
for 48 h (corresponding to the area shown in the FESEM image b1), the insets are
FESEM micrographs of O1 after solvent extraction at low-magnification (b1) and high-
magnification (b2).
This indicates significant difference in the interfacial interaction
between the two samples. As expected, the dispersion states of F2
and O1 in hot xylene at 135 ꢀC were different. The system of F2
showed a homogeneous dispersion in hot xylene, confirming that
the MMT sheets in the F2 could be stably dispersed in hot xylene.
This is because some PE chains were anchored to the MMT sheets
by copolymerization of ethylene with the copolymerizable vinyl on
the modifier molecule. In contrast, the MMTs in the O1 system had
deposited on the bottom because of no direct connection between
PE and MMT sheets. The difference in the dispersion states of F2
and O1 in hot xylene clearly confirms that the copolymerization
between F-MMT and ethylene should be carried out in the sample
of F2.
Rheological analysis further confirmed the above conclusion.
Generally the linear viscoelastic response of a nanocomposite melt
is strongly influenced by the factors, such as dispersion state of
particle, inter-particle and particleepolymer interactions. Fig. 12
shows the plots of storage modulus (G0) and complex viscosity
(j
h
*j) versus frequency (
u) for the PE/MMT nanocomposites. O1 and
HOMO show a nearly liner response in the plots of log j
h
*j vs log
u
because of the existence of long chain branch [31,32], which were
produced via the insertion of vinyl-terminated PE formed in situ
through b-hydride elimination or chain transfer to the monomer
[33e35]. In the case of F2, the presence of F-MMTs significantly
changed this behavior, especially in the low-frequency region. The
Table 2
The permeability coefficient of oxygen for the samples at 30 ꢀC.
G’ and j
h
*j of F2 at low frequencies were much higher than those of
O1. More importantly, a strong tendency of G0 to plateau and
Sample
HOMO
O1
F2
a significant j *j upturn at low-frequency regime was observed in
h
MMTs content (wt%)
0
8.2
8.0
0.95
0.15
the F2, which is widely accepted as the evidence of a network-like
structure and strong interfacial interaction [29,32,36e42].
PE is a typical polymer in the film application, but its barrier
property is poor. Adding clay into the PE matrix to form a PE/clay
Transmission rate (barrera)
6.15
1
2.02
0.33
b
Pn/Ph
a
1 barrer ¼ 10ꢂ10 cm cm3 cmꢂ2 s cmHg.
b
Ph: oxygen transmission rate of the HOMO sample; Pn: oxygen transmission rate
nanocomposite is
a promising method to reduce the gas
of the nanocomposite sample.

3424
C. Ren et al. / Polymer 51 (2010) 3416e3424
50525311), Key Project (20734006) and the Fund for Creative
Research Groups (no. 50921062).
References
[1] Jin YH, Park HJ, Im SS, Kwak SY, Kwak S. Macromol Rapid Commun
2002;23:135.
[2] Sun T, Garces JM. Adv Mater 2002;14:128.
[3] Bergman JS, Chen H, Giannelis EP, Thomas MG, Coates GW. Chem Commun
1999;21:2179.
[4] Lee DH, Kim HS, Yoon KB, Min KE, Seo KH, Noh SK. Sci Technol Adv Mater
2005;6:457.
[5] Wei L, Tang T, Huang B. J Polym Sci Part A: Polym Chem 2004;42:941.
[6] Liu C, Tang T, Wang D, Huang B.
2003;41:2187.
[7] Ray S, Galgali G, Lele A, Sivaram S.
2005;43:304.
J
Polym Sci Part A: Polym Chem
J Polym Sci Part A: Polym Chem
[8] Heinemann J, Reichert P, Thomann R, Mulhaupt R. Macromol Rapid Commun
1999;20:423.
[9] Tudor J, Willington L, O’Hare D, Royan B. Chem Commun 1996;17:2031.
[10] Huang Y, Yang K, Dong JY. Macromol Rapid Commun 2006;27:1278.
[11] Alexandre M, Dubois P, Sun T, Garces JM, Jerome R. Polymer 2002;43:2123.
[12] Du K, He AH, Liu X, Han CC. Macromol Rapid Commun 2007;28:2294.
[13] Vaia RA, Giannelis EP. Macromolecules 1997;30:7990.
[14] Balazs AC, Singh C, Zhulina E, Lyatskaya YL. Acc Chem Res 1999;32:651.
[15] Barus S, Zanetti M, Lazzari M, Costa L. Polymer 2009;50:2595.
[16] Wang WS, Chen HS, Wu YW, Tsai TY, Chen-Yang YW. Polymer 2008;49:
4826.
[17] Wang ZM, Nakajima H, Manias E, Chung TC. Macromolecules 2003;36:8919.
[18] Shin SYA, Simon LC, Soares JBP, Scholz G. Polymer 2003;44:5317.
[19] Atiqullah M, Faiz M, Akhtar MN, Salim MA, Ahmed S, Khan JH. Surf Interface
Anal 1999;27:728.
Fig. 13. XRD profiles and TEM images of F1 and F2 samples after annealing at 170 ꢀC
for 20 min.
4. Conclusions
We demonstrated an efficient in situ intercalative polymerization
method for synthesizing exfoliated PE/MMT nanocomposites with
good structural stability. This was on the basis of using the multi-
functional ammonium modifier bringing carbonyl group and vinyl
group. The F-MMT, originated from modifying MMT by the multi-
functional modifier, could anchor more the metallocene catalyst
within the gallery of MMT due to the presence of carbonyl group in
the F-MMT. In this case, more PE chains could be formed in the
interlayer of F-MMT during the in situ intercalative polymerization,
which leads to the formation of more exfoliated structure in the
resultant products, even at high content of F-MMT. The as-produced
PE/F-MMTs nanocomposites were composed of flower-like particles
[20] Grieken R, Martín C, Moreno J, Prieto O, Bravo JM. Macromol Symp
2007;259:174.
[21] Hammawa H, Wanke SE. Polym Int 2006;55:426.
[22] Galli P, Luciani L, Cecchin G. Angew Makromol Chem 1981;94:63.
[23] Burgentzlé D, Duchet J, Gérard JF, Jupin A, Fillon B.
2004;278:26.
J Colloid Interf Sci
[24] Russell KE, Hunter BK, Heyding RD. Polymer 1997;38:1409.
[25] Butler MF, Donald AM, Bras W, Mant GR, Derbyshire GE, Ryan AJ. Macro-
molecules 1995;28:6383.
[26] Ottani S, Wagner BE, Poter RS. Polym Commun 1990;13:370.
[27] Benn R, Fink G, Herrmann W, Müller T. Makromol Chem Rapid Commun
1992;13:321.
[28] Chiu F-C, Fu S-W, Chuang W-T, Sheu H-S. Polymer 2008;49:1015.
[29] Lee KM, Han CD. Macromolecules 2003;36:7165.
[30] Krishnamoorti R, Giannelis EP. Macromolecules 1997;30:4097.
[31] Lohse DJ, Milner ST, Fetters LJ, Xenidou M, Hadjichristidis N, Mendelson RA,
et al. Macromolecules 2002;35:3066.
[32] Ye Z, AlObaidi F, Zhu S. Ind Eng Chem Res 2004;43:2860.
[33] Wang WJ, Yan D, Zhu S, Hamielec AE. Macromolecules 1998;31:8677.
[34] Kolodka E, Wang WJ, Zhu S, Hamielec AE. J Appl Polym Sci 2004;92:307.
[35] Stapleton RA, Chai J, Nuanthanom A, Flisak Z, Nele M, Ziegler T, et al.
Macromolecules 2007;40:2993.
[36] Goharpey F, Nazockdast H, Katbab AA. Polym Eng Sci 2005;45:84.
[37] Tang CY, Xiang LX, Su JX, Wang K, Yang CY, Zhang Q, et al. J Phys Chem B
2008;112:3876.
with a diameter of about 5 mm. The formation of a thermal stable
monoclinic phase in PE/F-MMT nanocomposites probably resulted
from the stress formed during the polymerization and the copoly-
merization of the multifunctional modifier with ethylene. Moreover,
some of PE chains in PE/F-MMT nanocomposites were chemically
linked to the surface of F-MMTsheets through the vinyl group of the
multifunctional modifier during ethylene polymerization. Therefore
there was strong interaction between MMT sheets and polymer
matrix, which could stabilize the microstructure of the nano-
composites during processing at high-temperature. We believed
that the above method will shed a light on the further exploration of
preparing polymer/MMT nanocomposites with exfoliated structure
and good structural stability, especially with special aggregate states
of nascent particles, such as flower-like particles.
[38] Kashiwagi T, Du FM, Douglas JF, Winey KI, Harris RH, Shields JR. Nat Mater
2005;4:928.
[39] Kharchenko SB, Douglas JF, Obrzut J, Grulke EA, Milger KB. Nat Mater
2004;3:564.
[40] Du F, Scogna RC, Zhou W, Brand S, Fischer JE, Winey KI. Macromolecules
2004;37:9048.
Acknowledgements
[41] Cipriano BH, Kota AK, Gershon AL, Laskowski CJ, Kashiwagi T, Bruck HA, et al.
Polymer 2008;49:4846.
We thank the financial supports from the National Natural
Science Foundation of China for the Outstanding Youth Project (no.
[42] Filippone G, Dintcheva NT, Acierno D, La Mantia EP. Polymer 2008;49:1312.