Mechanochemical route to graphene-supported iron catalysts for olefin polymerization and in situ formation of carbon/polyolefin nanocomposites

Article abstract of DOI:10.1021/ma501602j

A facile one-step mechanochemical process converts graphite into highly active graphene-supported iron catalysts for ethylene polymerization and the in situ formation of graphene/polyethylene nanocomposites. Key feature is the dry grinding of graphite in

Full text of DOI:10.1021/ma501602j

Article
pubs.acs.org/Macromolecules
Mechanochemical Route to Graphene-Supported Iron Catalysts for
Olefin Polymerization and in Situ Formation of Carbon/Polyolefin
Nanocomposites
F. Beckert,†,‡,§ S. Bodendorfer,†,‡,§ W. Zhang,† R. Thomann,‡ and R. Mulhaupt*,†,‡
̈
†Institute for Macromolecular Chemistry, Albert-Ludwigs-University of Freiburg, Stefan-Meier-Straße 31, D-79104 Freiburg,
Germany
‡Freiburg Materials Research Center FMF, Stefan-Meier-Straße 21, D-79104 Freiburg, Germany
ABSTRACT: A facile one-step mechanochemical process
converts graphite into highly active graphene-supported iron
catalysts for ethylene polymerization and the in situ formation
of graphene/polyethylene nanocomposites. Key feature is the
dry grinding of graphite in a steel ball mill under carbon
dioxide pressure, affording high surface area edge-carboxylated
graphene accompanied by simultaneous immobilization of
Fe2+, formed by iron abrasion and electron transfer reaction. In
contrast, as evidenced by Mossbauer spectroscopy, grinding
̈
graphite in the absence of carbon dioxide under nitrogen or
argon pressure produces predominantly Fe0 together with Fe3+ supported on nitrogen-functionalized graphene or micronized
graphite, respectively. On addition of bisiminopyridine (BIP) and activation with methylaluminoxane (MAO), only the Fe2+
catalyst supported on edge-functionalized graphene (Fe@MG-CO2) polymerizes ethylene in high yields, producing polyethylene
with a molar mass of 180 kg/mol and a polydispersity of 6.1. According to the transmission electron microscopic analysis of
polyethylene morphology, functionalized graphene with low aspect ratio is uniformly dispersed in the polyethylene matrix.
Hence, this mechanochemical catalyst preparation enables the fabrication of graphene/polyolefin nanocomposites with high
carbon content by polymerization filling using cost-effective graphite as raw material without requiring tedious and expensive
graphite functionalization in separate steps.
wet grinding process employs liquid media with matched
polarity,18−20 frequently assisted by the addition of surfactants
such as SDS18 or melamine.21 In a dry grinding process, Jeon et
al.22 reported on the mechanochemical formation of edge-
carboxylated graphene produced by ball milling of graphite
together with dry ice. According to their hypothesis, the in situ
formed dangling bonds react with CO2 and incorporate
carboxylic acid groups allocated exclusively at the graphene
edges. This is in contrast to functionalized graphene produced
from graphite oxide, which decomposes to form nanoporous
graphene with much higher content of structural defects. In a
recent advance, the same group published the direct
mechanochemical nitrogen fixation by ball-milling of graphite
under N2 atmosphere.23 In all mechanochemical syntheses,
metal impurities originating from steel abrasion must be
removed in a further step, using either magnetic forces24 or
washing with hydrochloric acid.25 No attempts have been made
to exploit the mechanical immobilization of iron complexes and
nanoparticles produced by abrasion during dry grinding. Iron
catalysts are attractive in numerous reactions, including growth
of carbon nanotubes,26 direct C−H transformation,27 Fischer−
INTRODUCTION
■
During the past decade, the development of graphene and
graphene-based polymeric composite materials is attracting
considerable attention in academia and industry, stimulated by
the pioneering advances of Geim and Novoselov.1 As a single
carbon layer of the graphite lattice and two-dimensional carbon
macromolecule, graphene exhibits ultrahigh modulus,2 optical
transparency combined with electrical conductivity3 and a large
specific surface area,4 which enables the immobilization of
metal nanoparticles and the use as support material for
manifold reactions.5−8 However, the availability of defect-free
graphene is still limited, owing to its expensive synthesis such as
chemical vapor deposition (CVD)9,10 or epitaxial growth.11 In a
more cost-effective preparation, functionalized graphene (FG)
is produced by the oxidation of graphite followed by chemical12
or thermal13 reduction. By tuning the reaction parameters, FG
materials are tailored regarding their functional group density,
oxygen content as well as the surface area and the electrical
properties. However, graphite intercalation, oxidation and
subsequent reduction requires the use of strong acids,14,15
hazardous oxidizing agents and frequently also toxic reducing
agents such as hydrazine.16,17 In a much more cost- and time-
efficient route, mechanochemical synthesis has been introduced
in order to achieve both mechanical delamination and
functionalization in a facile one-step grinding process. The
Received: August 5, 2014
Revised: September 24, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ma501602j | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Tropsch synthesis,28 and epoxidation of alkenes29 as well as
precipitated in methanol (600 mL) with BHT (0.01 wt %) as
stabilizer. The polymer was filtered and dried in vacuum (10−2 mbar)
at 60 °C for 6 h. Table 1 summarizes the experimental conditions of
Fe-catalyzed polymerizations.
ethylene polymerization.30
Herein we report on the facile mechanochemical synthesis of
a highly active iron ethylene polymerization catalysts supported
on edge-carboxylated graphene. This dry grinding process
exploits graphite as raw material. Whereas most immobilization
strategies require several steps to immobilize iron on
functionalized graphene and other support materials,31−33 this
mechanochemical approach combines formation of edge-
carboxylated graphene support and in situ immobilization of
iron in a single step without requiring either solvent addition or
other chemical functionalization. Furthermore, we exploit this
mechanical process for producing in situ carbon/polyolefin
nanocomposites by polymerization filling technique.
Table 1. Experimental Conditions of the Fe-Catalyzed
Ethylene Polymerizations
a
m(MG-X)
[mg]
n(Fe)
n(BIP)
[μmol]
Al:Fe
[mol/mol]
[μmol]/[wt %]
FeBIP
−
250
4.47
4.47
44.7
100:1
100:1
Fe@MG-N2-
Ar
44.7/1
Fe@MG-CO2-
atm
250
250
178/4
44.6/1
200
100:1
100:1
Fe@MG-CO2-
Ar
44.7
EXPERIMENTAL SECTION
Materials. Natural graphite (KFL 99.5 from AMG Mining former
a
■
Determined by EDX measurements.
Kropfmuhl) was dried at 60 °C under vacuum (10 mbar) for 48 h
̈
prior to use. CO2, N2, and argon were received from Air Liquide and
used without further purification. All polymerization reactions were
carried out under dry argon atmosphere using standard Schlenk
techniques and a glovebox. Toluene (anhydrous) and n-heptane
(anhydrous) were purchased from Sigma-Aldrich. The solvents were
purified using a Vacuum Atmospheres Co. solvent purifier. MAO,
purchased from Crompton, with Al content of 10 wt % in toluene was
stored under argon atmosphere. Ethylene was purchased from Air
Liquide and was used without any further purification.
Instrumentation. Transmission electron microscopy (TEM) was
performed using a Zeiss/LEO 912 W at 120 kV. The samples were
cryo-microtomed or directly collected from dispersions on Cu grids.
Scanning electron microscopy (SEM) images were obtained with a
Quanta 250 FEG from FEI using backscattered mode. The
accelerating voltage was set to 20 kV. Atomic force microscopy
(AFM) measurements were performed on MultiMode AFM with a
Nanoscope IIIa controller (Veeco DI Instruments), using tapping
mode and Si cantilevers of supersharp type with about 2−10 nm radius
of curvature and 160 kHz resonance frequency. FT-IR spectra were
measured, using KBr specimen containing the sample. With a Vektor
22 from Bruker 32 scans with a resolution of 2 cm−1 were recorded. As
background the spectrum of a pristine KBr disc was used. The C, H,
and N ratio was determined using elemental analysis with a VarioEL
from Elementaranalysensysteme GmbH. The Fe, O, and Zr ratio was
determined with energy-dispersive X-ray spectroscopy (EDX) using an
One milling reactor was custom-made of stainless steel (1.4301 −
X2CrNi18−10; V = 370 cm3), whereas the zirconia reactor (V = 1150
cm3) was received from Retsch and equipped with a custom-made lid.
Preferably, the balls (50 pieces) and the mill were made of the same
materials. The samples were milled using a planetary ball mill PM 100
from Retsch.
Synthesis of MG-X (Typical Procedure). Graphite (5.0 g) was
added to a dried mill chamber (48 h at 60 °C, 10 mbar) and evacuated
(0.1 mbar) for 15 min. Afterward the milling reactor was pressurized
with CO2, N2, or Argon (7 bar) and milling was performed at 250 rpm
for the duration of 48 or 96 h. Specific samples were stored under inert
argon atmosphere, since exposure to air was accompanied by violent
sparkling and formation of red glowing embers.
Inca x-act from Oxford Instruments operating at 20 kV. Mossbauer
̈
spectra were recorded with a 57Co source in a Rh matrix using an
alternating acceleration Wissel Mossbauer spectrometer operated in
̈
the transmission mode and was equipped with a Janis closed-cycle
helium cryostat. Isomer shifts were recorded relative to iron metal at
ambient temperature.
Synthesis of the Bisiminopyridine Ligand (BIP). 2,6-
Diacetylpyridine (5.0 g, 31 mmol) and 2,6-dimethyaniline (8.1 mL,
65 mmol) were suspended in toluene (100 mL). A catalytic amount of
p-toluenesulfonic acid was added to the suspension. The reaction
mixture was stirred at 140 °C for 2 h with adapted water separator.
Afterward the solvent was removed at reduced pressure, the solid
residue was recrystallized from ethyl acetate (2 × 2 mL), and washed
with methanol (3 × 5 mL). The product was diluted in benzene and
extracted by freeze-drying.
Molecular weight determination was performed with a PL-200
chromatograph (Polymer Laboratories) using differential refractive
index (DRI). Measurements were performed at 150 °C with three
PLGel Olexis columns, calibrated with 12 polystyrene samples with a
narrow molecular weight distribution. 1,2,4-Trichlorobenzene (Merck)
was used as solvent, stabilized with 2,6-di-tert-butyl-(4-methylphenol)
(0.2 wt %, Aldrich). A flow rate of 1.0 mL/min was used. Thermal
properties were determined with differential scanning calorimetry
(DSC) using a Pyris 1 (PerkinElmer). The samples were heated from
room temperature to 260 °C with a heating rate of 20 K/min. Then
cycles were performed using a heating rate of 10 and 5 K/min. For
analysis of crystallization behavior a heating rate of 10 K/min was
used.
Polymers were stabilized using 0.5 wt % of a Irgafos 168/Irganox
1010 (1/4) blend. Melt compounding was performed using a DSM
twin screw micro compounder (XPlore) at 190 °C for 2 min.
Specimens were prepared by DSM micro injection molding (XPlore)
at 190 °C. Young’s modulus and elongation at break were measured by
stress−strain test (Zwick Z005) according to EN ISO 527−2. Results
are the average out of three measurements.
1H NMR (300 MHz, CDCl3/TMS): δ = 2.05 (s, 1′-H3, 2′-H3), 2.24
(s, 3′-H3), 6.94 (dd, J12−11 = J12−13 = 7.5, 12-H), 7.08 (d, J11−12 = J13−12
= 7.6, 11-H, 13-H), 7.91 (dd, J4,3 = J4,5 = 7.4, 4-H), 8.48 (d, J4,3 = J4,5
=
7.4, 3-H, 5-H).
Ethylene Polymerization. Reactions were carried out in a 600
mL Buchi stainless steel autoclave equipped with a mechanical stirrer
̈
and a software interface to analyze polymerization kinetics.
The MG-X support was preheated at 40 °C for 1 h. Then it was
dispersed in toluene (20 mL) and sonicated for 10 min. The
bisiminopyridine ligand (BIP) was dissolved in toluene (10 mL) and
added to the MG-X dispersion, followed by sonication for 60 min.
After the addition of MAO, the mixture was stirred for 5 min and
sonicated for further 10 min. The activated dispersion was washed with
fresh toluene to remove excess of MAO and BIP. The reactor was
charged with n-heptane (567 mL) and 50% of the needed MAO
amount as scavenger. During the polymerization the temperature was
kept at 40 °C with an ethylene pressure of 5 bar and a polymerization
time of 120 min while applying a stirring speed of 400 s−1. The
reaction was quenched by venting the vessel. The reaction mixture was
RESULTS
■
Pioneered by Brookhart et al., the methylalumoxane (MAO)
activation of bisiminopyridine iron complexes is well-known to
afford highly active single-site catalysts for olefin polymer-
ization.34 Such catalysts systems were immobilized on various
nanofillers in order to produce in situ nanocomposites with
B
dx.doi.org/10.1021/ma501602j | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
uniform dispersion of the nanofiller in the polyolefin matrix.35
This in situ dispersion during polymerization, named polymer-
ization filling technique by Dubois, takes place in gas phase or
diluents with low viscosity. This is advantageous with respect to
melt compounding, since nanofillers are rather difficult to
disperse in highly viscous polymer melts. In the case of carbon/
polyolefin nanocomposites the high aspect ratio of carbon
nanofillers plays an important role with respect to achieving
matrix reinforcement and low percolation threshold, as
reflected by high electrical conductivity at low carbon
content.36,37 While micronized graphite with high carbon
content, mostly consisting of particles >10 μm, is extensively
used as filler since many years, little is known concerning the
use of graphene produced by mechanochemical processes.
During recent years, Jeon et al. introduced a very versatile
mechanochemical process for preparing edge-functionalized
graphene. They reported on dry ball milling of graphite in the
presence of dry ice in order to produce edge-selective
carboxylated graphene in a process which is suitable for scale-
up.22 Moreover, ball milling under nitrogen enables the
incorporation of nitrogen into graphene. XPS studies revealed
the presence of C−N and N−N, thus Jeon et al. proposed the
formation of 5- and 6-membered rings at the edges of the
graphene platelets.23 In contrast, milling under argon
atmosphere under identical conditions fails to incorporate
functional groups in the absence of oxygen. In the presence of
air, the resulting nanometer-scaled carbon is readily oxidized.
Typically, iron nanoparticles, resulting from steel abrasion, are
removed by washing with HCl. Here, we exploit steel abrasion
to prepare iron catalysts supported on carboxylated and
nitrogenated graphene, produced by grinding graphite in a
planetary ball mill under carbon dioxide or nitrogen pressure,
respectively. Hence, the simultaneous delamination of graphite
and immobilization of iron by steel abrasion is achieved. Instead
of adding dry ice, which contains water traces, we pressurized
the ball mill with carbon dioxide during grinding in order to
achieve edge-selective carboxylation (see Figure 1). Typically,
the gas pressure of carbon dioxide, nitrogen and argon varied
between 7 and 13 bar.
measured by means of elemental analysis (EA) and energy-
dispersive X-ray spectroscopy (EDX).
As it is apparent from Table 2, milling in the presence of
carbon dioxide leads to high oxygen content up to around 17
wt %, mainly arising from the formation of carboxylic groups.
The iron content is highly dependent on milling time, however
the Fe-content in MG-CO2-96-S is overestimated due to the
combustion of carbon after violent sparkling when exposed to
air. Under nitrogen pressure a mechanochemical nitrogen
fixation of 5−8 wt % takes place, as previously reported by Jeon
et al.23 The resulting oxygen content of ∼10 wt % is due to
oxidation when the milling reactor was opened in the presence
of air. In contrast, ball milling of graphite under argon produced
nonfunctionalized nanometer-scaled carbon in the absence of
air. Obviously, highly reactive dangling bonds, formed during
the milling process, are capable of reacting with carbon dioxide,
nitrogen and also oxygen when carbon is exposed to air after
milling.
The incorporation of functional groups was verified by means
of IR spectroscopy. The FT-IR spectrum of the pristine
graphite (Figure 2a) exhibits only a broad band at 3400 cm−1
and a weak signal at 1620 cm−1 indicating adsorbed water. The
carboxylated MG-CO2 shows the characteristic bands at 1710
cm−1 corresponding to the CO stretching vibration, at 1585
cm−1 representing the CO2 asymmetric stretching vibration
−
and at 1225 typical for the C−O−C mode. In addition, the
products milled under argon and nitrogen (see Figure 2, traces
b and c) showed C−H stretching vibration modes at 2920 and
2840 cm−1. The peak at 1635 cm−1, representing isolated CC
stretching modes, might be due to the extensive destruction of
the aromatic π-system during the milling process.
Since all products were easily dispersible in most common
solvents, TEM was used to analyze the particle morphology.
After 12 h milling time the sheet like structure of the pristine
graphite was partially maintained, as could be seen by the large
and thin platelets. After 48 h of grinding, the particle size, as
well as the platelet structure vanished. Only spherical
nanoparticles were observed, as is apparent from the detailed
enlargement, with average diameter of a few nanometers.
Moreover, also the analysis of the AFM height profile
confirmed that after prolonged grinding most nanoparticles
exhibited cubic and spherical morphologies, whereas graphene-
like flat structures were absent. Obviously, prolonged grinding
affords high functionality at the expense of severe losses of the
aspect ratio, as evidenced by the disappearance of the
nanoplatelet structures especially after 96 h of milling. This is
paralleled by an increase of the specific surface area from <10
m2/g in graphite up to 350 m2/g in MG-CO2 as measured by
the BET method.
Figure 1. Edge-carboxylated graphene by dry grinding under CO2
pressure.
Only when grinding was performed in stainless steel
planetary ball mills in the presence of carbon dioxide and
nitrogen, up to ∼30 wt % iron was detected after 96 h in the
nanocarbon, whereas under argon and in ceramic ball mills no
iron was incorporated. Yet, grinding with zirconia ball mills
afforded only marginal amounts of zirconium impurities
resulting from abrasion of the zirconia. While Jeon et al.
carefully removed iron by treatment and washing with HCl22
we used such iron “impurities” on nanocarbon to prepare
carbon-supported iron catalysts for ethylene polymerization. It
should be noted, that the electron microscopic investigation
(see Figure 3) did not reveal the presence of iron nanoparticles.
In order to examine the role of metal abrasion, we used
planetary ball mills with two different milling reactors and balls,
both of which are made of either types of stainless steel or
yttrium-stabilized zirconia ceramic. Typically, the milling
reactor was charged with different amounts of dried graphite,
evacuated to remove traces of oxygen and moisture, and then
pressurized with oxygen-free dry carbon dioxide, nitrogen and
argon. After grinding at 250 rpm for 48 h, the reactor was
discharged under argon in order to prevent reaction with
moisture and air. As observed by Jeon, when opening under air,
violent sparkling occurred. The experimental results are
summarized in Table 2, listing the elemental composition and
metal content of the resulting nanocarbon materials, as
Therefore, we used Mossbauer spectroscopy to identify iron
̈
C
dx.doi.org/10.1021/ma501602j | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Table 2. Elemental Composition of Milled Graphite (250 rpm, 48 h, 7−13 bar gas Pressure)
a
a
a
b
b
b
sample
gas
reactor
C
[wt %]
H
[wt %]
N
[wt %]
O
[wt %]
Fe [wt %]
Zr [wt %]
MG-CO2-48-S
MG-CO2-96-S
CO2
CO2
CO2
N2
steel
76.6
27.8
79.2
83.0
79.7
87.8
88.8
96.4
0.8
0.9
0.8
0.5
0.3
0.4
−
−
−
−
8.1
5.3
0.4
−
17.8
27.6
17.3
11.5
10.4
−
0.9
29.1
−
−
−
0.8
d
steel
MG-CO2-48-C
MG-N2-48-S
MG-N2-48-C
MG-Ar-48-S
MG-Ar-48-C
ceramic
steel
0.3
−
2.2
−
N2
ceramic
steel
−
1.1
−
Ar
Ar
ceramic
10.2
0.8
c
graphite
−
b
−
0.5
−
−
−
−
a
c
d
From elemental analysis. From EDX. Pristine graphite prior to milling. Violent sparkling after exposure to air; carbon combustion leads to
underestimation of C-content and overestimation of Fe-content.
of ferric sites is well below 1 mm/s.38 In contrast, grinding
under nitrogen pressure produced iron-containing nitrogenated
graphene, which does not stabilize Fe2+ by complexation. In
fact, according to Figure 4b, Fe@MG-N2 contained predom-
inantly Fe0. Although more research is needed to clarify the
mechanisms accounting for the Fe2+ formation on edge-
carboxylated graphene, it is highly likely that electron transfer
reactions occur either with radicals of graphene formed by
prolonged grinding or with carboxyl radicals, formed by an
addition mechanism as proposed by Jeon. However, the
absence of any Fe2+ species in case of milling under nitrogen
pressure indicates an electron transfer reaction only possible by
the presence of MG-CO2 in accord with observations by
Jeon.22 It is well-known, that carboxylates stabilize iron(II)39 by
complexation. However, when exposed to oxygen, ferrous sites
readily oxidize to afford the corresponding ferric sites.
Using polymerization conditions as previously reported in
more detail,40 the Fe@MG-X were dispersed in n-heptane and
activated by adding the bisiminopyridine (BIP) ligand, followed
by activation with methylaluminoxane (MAO) after washing.
Ethylene polymerization was carried out in n-heptane at 5 bar
ethylene pressure and 40 °C, using an Al/Fe molar ratio of 100.
The polymerization conditions and results are listed in Tables 1
and 3.
Figure 2. FT-IR spectra (KBr specimen) of the (a) pristine graphite,
(b) MG-CO2, (c) MG-Ar, and (d) MG-N2.
and its oxidation state obtained after prolonged grinding of
graphite.
According to Mossbauer spectroscopy, the Fe@MG-CO
̈
2
sample, prepared by grinding under carbon dioxide pressure
and stored under argon, contained predominantly Fe2+ together
with Fe0 (see Figure 4a). The characteristic new signal at
chemical shift CS = 1.1 mm/s was only present if milled under
carbon dioxide and disappeared if exposed to air. A chemical
shift of CS > 1 mm/s is typical for ferrous sites, whereas the CS
Figure 3. Micrographs of MG-CO2: TEM images were taken from acetone dispersions: (a) ball milling for 12 h and (b) ball milling for 48 h. SEM
image of a sample milled for 48 h (c) and 96 h (e). AFM measurement and height profile of MG-CO2-48 h (d).
D
dx.doi.org/10.1021/ma501602j | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 4. Fe−Mossbauer spectra recorded at 80 K of (a) Fe@MG-CO stored under argon and (b) Fe@MG-N .
̈
2
2
Table 3. Catalyst Composition and Polymer Properties
b
a
support Fe@MG-
surface area [m2/g]
n(Fe) [μmol]
yield [g]
C-content [wt %]
molar mass (Mw) [kg/mol]
Mw/Mn
Tm [°C]
FeBIP
−
223
271
273
n.d.
n.d.
4.47
44.6
128
0
−
−
2.2
0.8
0.2
5.0
165
−
181
177
283
80
16
−
6.1
8.3
6.1
6.3
132
−
135
134
136
131
N2-Ar
CO2-atm
CO2-Ar-48
CO2-Ar-96 (0.2)
178
11.4
30.5
117
49.4
44.6
250
c
CO2-Ar-96 (5.0)
2500
a
b
Sum of Fe0, FeII, and FeIII calculated by EDX measurement. Polymerization parameters: m(support) = 250 mg, Vpol = 600 mL, tpol = 120 min, Al/
c
Fe = 100, ppol = 5 bar, Tpol = 40 °C. Yield-controlled polymerization for definite filler content; tPol = 10 min.
From Table 3 it is apparent that under the chosen conditions
only the homogeneous catalyst FeBIP and the iron catalysts
supported on edge-carboxylated graphene are active in ethylene
polymerization, whereas Fe/C produced in the presence of
nitrogen (Fe@MG-N2) and argon are inactive. This is in accord
activity. Interestingly, ethylene polymerization on Fe@MG-
CO2-Ar-96 produces the same amount of polyethylene as the
polymerization with neat FeBIP. Figure 5 shows kinetics of the
supported catalysts. It is obvious that milling graphite in a steel
mill under CO2 and storing it in the absence of oxygen under
argon affords highly active catalysts. The performance of these
catalysts leads to a yield that is in the same range as
homogeneous MAO/FeBIP. The catalyst activities markedly
increase with increasing milling time, showing that the grinding
duration plays an important role. Milling under nitrogen or
milling under argon or subsequent exposure to air drastically
lowers polymerization activity. Obviously, oxidation of Fe2+ to
Fe3+ significantly reduces the content of active sites. As is
apparent from Table 3, the Mw of 181 kg/mol as well as the size
distribution (PDI = 6.1) are in good agreement with the
reported results of PE polymerizations realized with MAO/
FeBIP and MAO-activated iron catalysts supported on silica.40
Fe@MG-CO2-96h (5.0) shows a lower molar mass which can
be a result of the short polymerization time of 10 min. The
excess of Fe in the reaction mixture attributed to a lower Mn
(10 kg/mol) as a consequence of a higher number of catalytic
centers. Since the Fe@MG-CO2 catalysts contain substantial
amounts of Fe0, the resulting carbon/Fe/polyethylenes are
magnetic. It should be noted that MG-CO2−Ar-48, pretreated
with MAO, can also be used as support for other single site
catalysts such as ZrCp2Cl2.
with the results of the Mossbauer spectroscopic investigation,
̈
indicating the presence of Fe2+ exclusively in the case of Fe@
MG-CO2. It should be considered, that the Fe content given in
Table 3 represents the sum of Fe0, FeII, and FeIII species. It was
not possible to quantitatively determine the content of active
FeII species, thus a relationship between Fe content and yield
(molar mass) is not given at this point. From the polymer-
ization kinetics displayed in Figure 5 it is apparent that
exposure of Fe@MG-CO2 to air, accompanied by massive
oxidation of ferrous to ferric sites, drastically impairs catalysts
The carbon/polyethylene nanocomposites produced by
ethylene polymerization on MAO/BIP-activated Fe@MG-
CO2 were melt processed by means of twin-screw mini-
extrusion and subsequent micro injection molding. According
to the transmission electron images (TEM) of PE containing
0.5 wt % Fe@MG-CO2 (see Figure 6), the carbon particle
diameter varies from a few nanometers to 500 nm in size. This
is not surprising because the broad particle size distribution is
identical to that of the milled carbon displayed in Figure 3,
Figure 5. Online polymerization kinetics of ethylene polymerization
on MAO-activated Fe@MG-X/BIP catalysts as compared to the
homogeneous FeBIP/MAO catalysts; polymerization parameters: tPol
= 10 min, Al/Fe = 100, ppol = 5 bar, Tpol = 40 °C.
E
dx.doi.org/10.1021/ma501602j | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
polyolefin nanocomposites prepared by polymerization filling
technique. Compared to graphene/polyethylene composites
prepared by conventional melt compounding, elongation at
break could be maintained, whereas Zheng et al reported on
massive losses of elongation at break.41
CONCLUSION
■
We have developed a facile one-step mechanochemical process
for the synthesis of iron polymerization catalysts supported on
functionalized multilayer graphene for the preparation of
carbon/polyolefin nanocomposites derived directly from graph-
ite. Key feature is formation of an edge-carboxylated graphene
as high-surface-area carbon support by chemical carbon dioxide
fixation during grinding of graphite under carbon dioxide
pressure. In this mechanochemical process, the formation of
edge-selective carboxylated graphene is accompanied by the
simultaneous immobilization of Fe2+, resulting from abrasion of
steel and electron transfer during milling. Unlike conventional
graphene-supported polymerization catalysts, neither tedious
multistep catalyst preparation nor the use of graphite oxide as
intermediate is required. Ethylene polymerization takes place
when Fe@MG-CO2 is activated by the subsequent addition of
Figure 6. TEM images of a polyethylene containing 0.5 wt % Fe@
MG-CO2-48h.
parts c and e. Some agglomerates are present in the polymer
composite. This is induced by the morphology of the pristine
MG-CO2 showing agglomerates and multilayered particles as a
result of the milling process. The carbon/polyethylene
nanocomposites were characterized by means of stress−strain
measurements in order to determine their Young’s modulus
and elongation at break. Figure 7 compares the mechanical
BIP and MAO. According to Mossbauer spectroscopy, the
̈
presence of Fe2+ is crucial for achieving activity in ethylene
polymerization. When Fe@MG-CO2 is exposed to air, ferrous
sites are converted into inactive ferric sites, thus accounting for
drastically lowering catalyst activities. Polymerization activity
increases with increasing milling time, which is paralleled by
increasing the amount of immobilized iron and carboxylate
groups. Further investigations showed, that the overall Fe
content is reliably controllable by milling time and CO2
pressure. This mechanochemical process for producing nano-
carbon-supported iron catalysts is an attractive route to carbon/
polyolefin nanocomposites prepared by the polymerization
filling technique. Owing to the drastic reduction of graphene
aspect ratio during prolonged grinding, the matrix reinforce-
ment of such edge-carboxylated graphene is considerably lower
with respect to that of graphene/polyolefin nanocomposites
based upon thermally reduced graphite oxide. However, this
mechanochemical route to carbon/polyolefin compounds is
cost-effective and enables the incorporation of much larger
carbon nanoparticle content without impairing processing. This
offers new opportunities for designing carbon/polyolefin
composite materials with high electrical and thermal con-
ductivity for potential applications in lightweight engineering.
AUTHOR INFORMATION
■
Corresponding Author
Figure 7. Young’s modulus and elongation at break (left) of carbon/
polyethylene composites (a) FeBIP, (b) Fe@MG-CO2-Ar-48h (0.8 wt
%), (c) Fe@MG-CO2-Ar-96h (0.2 wt %), and (c) Fe@MG-CO2-Ar-
96h (5.0 wt %) and stress/strain−curves of the samples (right).
*(R.M.) E-mail: rolf.muelhaupt@makro.uni-freiburg.de.
Author Contributions
§Authors contributed equally.
Notes
properties of the polyethylene produced by homogeneous
FeBIP/MAO and by Fe@MG-CO2/BIP/MAO polymerization,
respectively. Obviously, the incorporation of Fe@MG-CO2
accounts for markedly improved matrix reinforcement at
carbon content of 5 wt % (Figure 7d). Thus, the mechanical
properties are not related to the Fe content, instead the content
of carbon based filler affect the mechanical properties. This is in
good agreement with the performance of other graphene/
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors gratefully acknowledge the financial support by the
European Commission within the Graphene Flagship (GAN
604391). The authors thank A. Warmbold and J. Eckerle for
their support doing the experimental work.
F
dx.doi.org/10.1021/ma501602j | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
(37) Fim, F. D.; Basso, N. R. S.; Graebin, A. P.; Azambuja, D. S.;
Galland, G. B. J. Appl. Polym. Sci. 2013, 128, 2630.
(38) Williams, A. G. B.; Scherer, M. M. Environ. Sci. Technol. 2004,
38, 4782.
(39) Lippard, S. J. Philos. Trans. R. Soc. A: Math., Phys. Eng. Sci. 2005,
363, 861.
REFERENCES
■
(1) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang,
Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Science 2004, 306,
666.
(2) Lee, C.; Wei, X.; Kysar, J. W.; Hone, J. Science 2008, 321, 385.
(3) Kim, K. S.; Zhao, Y.; Jang, H.; Lee, S. Y.; Kim, J. M.; Kim, K. S.;
Ahn, J.-H.; Kim, P.; Choi, J.-Y.; Hong, B. H. Nature 2009, 457, 706.
(4) Cranford, S. W.; Buehler, M. J. Phys. Rev. B 2011, 84, 205451.
(5) Scheuermann, G. M.; Rumi, L.; Steurer, P.; Bannwarth, W.;
(40) Kurek, A.; Mark, S.; Enders, M.; Kristen, M. O.; Mulhaupt, R.
̈
Macromol. Rapid Commun. 2010, 31, 1359.
(41) Zheng, W.; Lu, X.; Wong, S.-C. J. Appl. Polym. Sci. 2004, 91,
2781.
Mulhaupt, R. J. Am. Chem. Soc. 2009, 131, 8262.
(6) Wan, Q.; Liu, Y.; Wang, Z.; Wei, W.; Li, B.; Zou, J.; Yang, N.
̈
Electrochem. Commun. 2013, 29, 29.
(7) Sturzel, M.; Kempe, F.; Thomann, Y.; Mark, S.; Enders, M.;
̈
Mulhaupt, R. Macromolecules 2012, 45, 6878−6887.
̈
(8) Lim, D.-H.; Wilcox, J. J. Phys. Chem. C 2012, 116, 3653.
(9) Gomez De Arco, L.; Zhang, Y.; Schlenker, C. W.; Ryu, K.;
Thompson, M. E.; Zhou, C. ACS Nano 2010, 4, 2865.
(10) Wang, X.; You, H.; Liu, F.; Li, M.; Wan, L.; Li, S.; Li, Q.; Xu, Y.;
Tian, R.; Yu, Z.; Xiang, D.; Cheng, J. Chem. Vap. Deposition 2009, 15,
53.
(11) Avouris, P.; Dimitrakopoulos, C. Mater. Today 2012, 15, 86.
(12) Mauro, M.; Cipolletti, V.; Galimberti, M.; Longo, P.; Guerra, G.
J. Phys. Chem. C 2012, 116, 24809.
(13) Tolle, F. J.; Fabritius, M.; Mulhaupt, R. Adv. Funct. Mater. 2012,
̈
̈
22, 1136.
(14) Brodie, B. C. Philos. Trans. R. Soc. London 1859, 149, 249.
(15) Hummers, W. S.; Offeman, R. E. J. Am. Chem. Soc. 1958, 80,
1339.
(16) Stankovich, S.; Dikin, D. A.; Piner, R. D.; Kohlhaas, K. A.;
Kleinhammes, A.; Jia, Y.; Wu, Y.; Nguyen, S. T.; Ruoff, R. S. Carbon
2007, 45, 1558.
(17) Eda, G.; Fanchini, G.; Chhowalla, M. Nat. Nanotechnol. 2008, 3,
270.
(18) Knieke, C.; Berger, A.; Voigt, M.; Taylor, R. N. K.; Rohrl, J.;
̈
Peukert, W. Carbon 2010, 48, 3196.
(19) Zhao, W.; Fang, M.; Wu, F.; Wu, H.; Wang, L.; Chen, G. J.
Mater. Chem. 2010, 20, 5817.
(20) Wu, H.; Zhao, W.; Chen, G. J. Appl. Polym. Sci. 2011, 125, 3899.
(21) Leon, V.; Quintana, M.; Herrero, M. A.; Fierro, J. L. G.; Hoz, A.
d. l.; Prato, M.; Vazquez, E. Chem. Commun. 2011, 47, 10936.
(22) Jeon, I.-Y.; Shin, Y.-R.; Sohn, G.-J.; Choi, H.-J.; Bae, S.-Y.;
Mahmood, J.; Jung, S.-M.; Seo, J.-M.; Kim, M.-J.; Wook Chang, D.;
Dai, L.; Baek, J.-B. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 5588.
(23) Jeon, I.-Y.; Choi, H.-J.; Ju, M. J.; Choi, I. T.; Lim, K.; Ko, J.; Kim,
H. K.; Kim, J. C.; Lee, J.-J.; Shin, D.; Jung, S.-M.; Seo, J.-M.; Kim, M.-
J.; Park, N.; Dai, L.; Baek, J.-B. Sci. Rep. 2013, 3, 2260.
(24) Lee Jae Kap, L. Kyoung-Il, Vol. US2012043402 (A1), K. I. S.
Tech.:Korea, 2010.
(25) Welham, N. J.; Williams, J. S. Carbon 1998, 36, 1309.
(26) Anna, M.; Albert, G. N.; Esko, I. K. J. Phys.: Condens. Matter
2003, 15, S3011.
(27) Sun, C.-L.; Li, B.-J.; Shi, Z.-J. Chem. Rev. 2010, 111, 1293.
(28) Dry, M. E. Catal. Today 2002, 71, 227.
(29) Lane, B. S.; Burgess, K. Chem. Rev. 2003, 103, 2457.
(30) Britovsek, G. J. P.; Bruce, M.; Gibson, V. C.; Kimberley, B. S.;
Maddox, P. J.; Mastroianni, S.; McTavish, S. J.; Redshaw, C.; Solan, G.
A.; Stromberg, S.; White, A. J. P.; Williams, D. J. J. Am. Chem. Soc.
1999, 121, 8728.
(31) Janiak, C.; Rieger, B. Angew. Makromol. Chem. 1994, 215, 47.
(32) Schilling, M.; Bal, R.; Gorl, C.; Alt, H. G. Polymer 2007, 48,
̈
7461.
(33) Schmidt, R.; Welch, M. B.; Palackal, S. J.; Alt, H. G. J. Mol. Catal.
A: Chem. 2002, 179, 155.
(34) Small, B. L.; Brookhart, M. J. Am. Chem. Soc. 1998, 120, 7143.
(35) Alexandre, M.; Dubois, P.; Sun, T.; Garces, J. M.; Jero
Polymer 2002, 43, 2123.
̂
me, R.
(36) Quan, Y. N.; Lu, M.; Tian, M.; Yan, S. K.; Yu, Z. Z.; Zhang, L.
Q. J. Appl. Polym. Sci. 2013, 130, 680.
G
dx.doi.org/10.1021/ma501602j | Macromolecules XXXX, XXX, XXX−XXX