Exfoliation of sheet silicates by nitroxide mediated polymerization

Article abstract of DOI:10.2533/chimia.2008.799

A new cationic alkoxyamine initiator/regulator for the nitroxide-mediated polymerization of acrylates was prepared and intercalated by cation exchange into a montmorillonite-type natural and a synthetic layered silicate. Thermal activation of this alkoxyamine in the presence of n-butyl acrylate as monomer initiated its controlled radical polymerization, leading to an almost complete exfoliation of the silicate layers. The poly(n-butyl-acrylate) chains attached to the individual sheet silicate layers showed adjustable number average molecular weights (M_n) values between 1000 and 14000 g/mol and polydispersities PD (= M_w/M_n; M_w = weight average molecular weight) as low as 1.5. The poly(n-butyl acrylate) modified silicate nanoparticles obtained by this process have a polymer content between 41 and 89 wt.%, are non-agglomerated and stable in dispersion for >1 year without any sedimentation. They are easily re-dispersible in apolar solvents, in monomers and macromonomers which are used e.g. in coating formulations and in polymer solutions/melts, forming nanocomposites with novel properties. For example, a 20% reduction of the O₂ permeability of polyethylene films was observed at 5 wt% loading with the exfoliated sheet silicate. Schweizerische Chemische Gesellschaft.

Full text of DOI:10.2533/chimia.2008.799

POLYMERS AND COLLOIDS DIVISION
799
CHIMIA 2008, 62, No. 10
doi:10.2533/chimia.2008.799
Chimia 62 (2008) 799–804
© Schweizerische Chemische Gesellschaft
ISSN 0009–4293
Exfoliation of Sheet Silicates by Nitroxide
Mediated Polymerization
Andreas Mühlebach*, Peter Nesvadba*, François Rime, and Lucienne Bugnon
Abstract: A new cationic alkoxyamine initiator/regulator for the nitroxide-mediated polymerization of acrylates was
prepared and intercalated by cation exchange into a montmorillonite-type natural and a synthetic layered silicate.
Thermal activation of this alkoxyamine in the presence of n-butyl acrylate as monomer initiated its controlled radical
polymerization, leading to an almost complete exfoliation of the silicate layers. The poly(n-butyl-acrylate) chains
attached to the individual sheet silicate layers showed adjustable number average molecular weights (Mn) values
between 1000 and 14000 g/mol and polydispersities PD (= Mw/Mn; Mw = weight average molecular weight) as low
as 1.5. The poly(n-butyl acrylate) modified silicate nanoparticles obtained by this process have a polymer content
between 41 and 89 wt.%, are non-agglomerated and stable in dispersion for >1 year without any sedimentation.
They are easily re-dispersible in apolar solvents, in monomers and macromonomers which are used e.g. in coating
formulations and in polymer solutions/melts, forming nanocomposites with novel properties. For example, a 20%
reduction of the O2 permeability of polyethylene films was observed at 5 wt% loading with the exfoliated sheet
silicate.
Keywords: Alkoxyamines · Controlled/living radical polymerization · Exfoliation · Layered silicates ·
Nanocomposites
1. Introduction
availability, the anisometric particle shape tions between the silicate layers are mini-
with a high aspect ratio (typically >200), mized and no periodic stacking is observed.
Polymer/layered silicates (clay) nanocom- the unique layered crystal structure and the Additionally, an appropriate choice of the
posites, combining the characteristics of fact that the thickness of the individual lay- polymer provides a methodology to tailor
both materials at the nanoscale level, have ers is in the nanometer range (ca. 1 nm). the compatibility of the layered silicate
recently found considerable industrial and However, a prerequisite for optimal proper- with the matrix.
academic interest for improvement of di- ties of the nanocomposite is good disper-
In this work we used the semi-commer-
verse properties of a variety of polymers or sion and excellent spatial distribution of the cial montmorillonite-type layered silicate
coatings. For example, polymer/layered sil- silicate nanofiller in the polymeric matrix Nanofil EXM 588 and the commercial
icates nanocomposites were shown to pos- material. This approach requires deagglom-
synthetic sheet silicate Optigel SH. Both
sess enhanced mechanical strength, higher eration and delamination (or exfoliation) materials consist of negatively charged Si/
barriers to gas permeation, better electrical of the individual silicate layers which are Al-oxide layers and single charged Na+ and
[1–7]
insulation or lower flammability.
The
held together by electrostatic forces into K+ cations which are located in the galleries
[13,14]
property enhancements are attributed to stacked agglomerates in the original clay. between these layers
and are character-
the large interface areas between the sili- Moreover, a suitable functionalization of ized by a relatively high ion exchange ca-
cate platelets and the host polymer matrix. the delaminated hydrophilic clay layers is pacity and a large specific surface area. The
The special interest in layered (sheet) sili-
cates arises from their low price and good ting) between the hydrophilic silicate layers are prone to exchange with other cations in
and the – in general – hydrophobic poly- an equilibrium reaction. The driving force
mer matrix. To this purpose, various pre- of this reaction is the gain of solvation en-
needed to increase the compatibility (wet- Na+ and K+ ions between the silicate layers
treatments of the clay silicates, such as for thalpy of the liberated cations relative to the
example treatment of the clay with silane intercalated ones and therefore the cation
[8]
coupling agents or replacement of the exchange is best performed in water with
[15]
clay interlayer metal cations with hydro- amphiphilic organic cations. Thus, modi-
[9]
phobic organic cations were developed. fied montmorillonite-type sheet silicates
Yet another possibility consists of growing with interlayer distance of ca. 1.5–2.5 nm
the polymeric chain from the surface of the are readily accessible by intercalation with
layered silicate. To do this, the clay is in-
tercalated by suitable organic cations, car- Some time ago we synthesized and in-
organic cations (onium ions).
*Correspondence: Dr. A. Mühlebach; Dr. P. Nesvadba
Tel.: +41 61636 2427; +41 61636 2412
E-mail: andreas.muehlebach@cibasc.com;
peter.nesvadba@cibasc.com
Ciba, R&D Coating Effects and Group Research
Klybeckstrasse 141
rying a reactive group (initiator, monomer) tercalated symmetric dicationic peroxide
that can participate in the polymerization initiators into layered silicates and studied
and promote thus exfoliation and separation their potential to induce silicate exfoliation
[10–12]
[16]
CH-4002 Basel
of the clay layers.
Hence, the interac- by uncontrolled radical polymerization.

POLYMERS AND COLLOIDS DIVISION
800
CHIMIA 2008, 62, No. 10
However, initiator efficiency was very low.
In fact, we estimated that only one out of
500 initiator molecules is active in polym-
erization and GPC analysis of the polymer
showed a high molecular weight (Mn >105
g/mol) combined with very high polydis-
Monomer
Y
kp
R1
R2
ka
kd
R1
R2
O
N
n
CH
Y
*
O
N
n
+
*
Y
Y
Y
Y
Y
Dormant chain
Propagating radical
Nitroxide
persity (PD >5). Therefore, not surprising-
ly, the exfoliation was incomplete.
On the other hand, the controlled/living
radicalpolymerization(CRP)isanexcellent
Scheme 1. Nitroxide-mediated polymerization (NMP)
method to control the polymer molecular
[17]
weight and architecture and is – in con-
trast to other living polymerization methods
– tolerant to impurities and moisture which
are present in the naturally occurring sheet
N
O
silicates. The most versatile CRP meth-
ods are atom-transfer radical polymeriza-
tion (ATRP), reversible radical addition-
fragmentation chain transfer (RAFT) and
nitroxide-mediated radical polymerization
OEt
N
N
P
N
OEt
O
O
O
O
1
2
3
(NMP). The underlying principle of these
OH
O
CRP techniques is the reversible activation/
deactivation of the propagating radicals. In
the case of NMP it is the reversible combi-
OH
N
O
nation of propagating radicals with nitrox-
N
O
N
O
C6H5
[18,19]
ide radicals
(Scheme 1).
A great number of publications report
the use of all three CRP techniques to
grow polymer brushes from planar sub-
strates or nanoparticles including silica,
aluminium or iron oxides, quantum dots
4
5
6
Fig. 1. Nitroxides and alkoxyamines for NMP of acrylates
or gold colloids but only a few works ex-
[20,21]
nitroxides, 1,1-dimethylethyl 2-methyl-1-
ist on clay minerals.
Sogah and co-
detailed results of the exfoliation of a natu-
ral (montmorillonite-type) and a synthetic
sheet silicate with the alkoxyamine 9 and
[24]
phenylpropyl nitroxide (1, TIPNO)
or the
above-mentioned N-tert-butyl-N-[1-dieth-
ylphosphono-(2,2-dimethylpropyl)-nitrox-
workers showed in a preliminary commu-
[22]
nication that polystyrene chains could be
grown from the surface of montmorillonite
via NMP, using an intercalated cationic te-
tramethyl-piperidine derived alkoxyamine
n-butyl acrylate as monomer. The synthesis
[25]
ide (2, DEPN, SG-1). In our laboratories
of additional cationic alkoxyamines derived
from 3, 4 or 5 and the corresponding exfo-
liation experiments have been published in
were developed sterically highly hindered
[26]
[23]
piperazin-3-one, e.g. 3
and piperidine
NMP initiator. In a more recent paper
[32]
nitroxides and the related alkoxyamines,
they described the grafting of poly(styrene-
ε-caprolactone) diblock copolymers from
the clay layers using an initiator anchored
on the clay layers and capable of initiating
simultaneously the ring-opening polymer-
ization of ε-caprolactone and NMP of sty-
rene. Very recently, Bourgeat-Lami and co-
workers applied NMP using a quaternary
ammonium alkoxyamine initiator prepared
from N-tert-butyl-N-[1-diethylphosphono-
(2,2-dimethylpropyl)-nitroxide (2, Fig. 1)
to grow polystyrene chains with controlled
molecular weight and relatively low poly-
dispersities (PD >1.5) from the surface of
our recent patent application.
[27,28]
e.g. compounds 4-6.
The latter are now
used in the industrially applied NMP pro-
[29–31]
cess
to produce acrylate block-copo-
2. Experimental
lymers with controlled molecular weight,
block composition and low polydispersity.
Building on these developments, we used
the nitroxide 5 and transformed it into the
alkoxyamine8viaCu(i)-promotedcoupling
with the readily available (see Experimen-
tal Section) bromoamide 7. Quaternization
of 8 with benzylchloride afforded the de-
sired cationic alkoxyamine NMP-initiator
9 (Scheme 2). In this article we present
2.1. Materials and Instruments
Solvents, reagents and monomers are
all commercially available and were used
as received. Na+-montmorillonite (Nano-
fil EXM 588), a natural sheet silicate, was
obtained from Süd-Chemie AG, Germany,
as a greyish powder. The cation exchange
capacity was specified to be 0.61 mmol/g
and the interlayer distance is 0.98 nm. Op-
[20]
Laponite clay.
In this article, we report the, to the best
of our knowledge, first montmorillonite
exfoliation via NMP of n-butylacrylate us-
ing novel cationic alkoxyamine intercalated
into montmorillonite by cation exchange.
At the beginning of CRP only styrene
monomers could be polymerized in a con-
trolled way by NMP. In the past few years,
several new nitroxides, allowing the con-
trolled polymerization of other monomers,
most importantly of acrylates, acrylonitrile
or acrylamides, were introduced. Amongst
OH
OH
OH
CuCl
N
C6H5CH2Cl
Acetonitrile, r.t.
PMDETA
N
O
EtOAc, r.t.
O•
N
5
O
Br
O
N
H
N
N+
Cl-
O
N
H
O
N
H
N
9
7
8
the most significant (Fig. 1) are open chain Scheme 2. Synthesis of the cationic alkoxyamine 9

POLYMERS AND COLLOIDS DIVISION
801
CHIMIA 2008, 62, No. 10
tigel SH, a synthetic sheet silicate with a 2.3. 2-(2,6-Diethyl-4-hydroxy-2,3,6-
cation exchange capacity of 0.78 mmol/g, trimethyl-piperidine-1-yloxy)-N-
nitrogen purging cycles and then heated
with an oil bath under nitrogen to 140 °C for
was also obtained from Süd-Chemie as a (3-dimethylamino-propyl)-2-methyl- 7 h. The conversion was determined by 1H-
white powder. Centrifugation was done in propionamide (8)
100–180 ml glass vessels in an IEC Cen- To a solution of 2,6-diethyl-2,3,6-
tra GP8 centrifuge at approx. 2000 rpm trimethyl-piperidine-4-hydroxy-N-oxyl (5, and polydispersity was measured by GPC.
(corresponding to ca. 850 g). 1H- and 13C- 12.86 g, 60 mmol, prepared as described
8: Conversion 71%, Mn = 6150 (predict-
NMR (integration of residual monomer vs.
polymer signals) and the molecular weight
NMR spectra were measured on a Bruker in ref. [28]) in EtOAc (50 ml) was added ed Mn = 6390), PD = 1.21, 9: Conversion
Avance 300 NMR spectrometer and ref- under argon CuCl (11.9 g, 120 mmol) and 65%, Mn = 6270 (predicted Mn = 6010), PD
erenced to SiMe4 (δ in ppm, J in Hz). IR 2-bromo-N-(3-dimethylamino-propyl)-2- = 1.34.
spectra were taken on Nicolet Magna-IR methyl-propionamide (7, 16.5 g, 66 mmol).
750 spectrometer in KBr pill, MS spectra Then, N,N,N’,N’’,N’’-pentamethyl-diethyl-
on Finnigan SSQ 710 apparatus. Molecu- ene-triamine (PMDTA) (20.8 g, 120 mmol) Layered Silicate ‘Nanofil EXM 588’
lar weights (Mn, Mw) and polydispersity was added dropwise within 55 min while (Montmorillonite-type) (Scheme 3)
(PD = Mw/Mn) were measured by GPC in keeping the temperature at 35–38 °C. The 60 g (116.9 mmol, 1.2 equiv. relative to
THF on a PL Gel (Polymer Laboratories) mixture was then stirred 15 h at room tem- the ion-exchange capacity of the sheet sili-
guard and 2 × 5 µm mixed bed C columns perature under argon. The green suspension cate) 9 was dissolved in a mixture of 850
(40 °C) with a Waters 600E pump and was filtered and the filter cake was washed ml distilled water and 150 ml EtOH. 159.6
controller and a Waters 410 RI detector. with 100 ml ethyl acetate. The filtrate was g Nanofil EXM 588 (containing 12.6 wt.%
2.6. Intercalation of 9 into the
[33]
The values are relative to monodisperse
washed successively with 3 × 50 ml water, water ) was dispersed in this solution and
[34]
PS standards. Thermogravimetric analy-
ses (TGA) were performed under air on solution, 50 ml water, 50 ml saturated NaCl The dispersion was centrifuged for 1 h,
a Mettler Toledo TC15 TA instrument and solution, dried over MgSO4 and evaporat- the wet residue dispersed again in 500 ml
50 ml of a 1% aqueous EDTA-disodium salt stirred for 24 h at room temperature.
controller with a heating rate of 10 °C/ ed, finally at 0.01 mbar/50 °C to afford 23 g water/EtOH (85/15) and centrifuged. This
min. Elemental analyses were obtained (99.4%) of 8 as a slightly yellow thick oil. procedure was repeated two more times,
from the Ciba Service Center Elemen-
1H-NMR (300 MHz, CDCl3), mix- the last time with pure EtOH. The product
tal Analysis. Residual solvent molecules ture of diastereoisomers: 7.37–7.25 (bs, was dried in vacuo (3 d, 50 °C, 0.01 mbar).
were identified by 1H-NMR. Powder NH), 4.19–4.11 (m, CH(OH)), 3.60–3.10 Yield: 184 g of a grey powder.
X-ray measurements were performed (m, CH2), 2.35–2.25 (m, CH2), 2.23 (s,
Weight loss determined by TGA (50–
600 °C): 27.1 wt%, corresponding to 56.9
mmol 9 in 100 g intercalated sheet silicate.
Powder X-ray analysis gave a main reflec-
tion peak at 2Θ = 3.86° (Fig. 2b), the origi-
nal strong reflection peak at 2Θ = 7.1° (Fig.
2a) has disappeared.
on a Siemens D500 diffractometer with N(CH3)2), 2.10–0.79 (m, 31H).
monochromatic Cu-Kα radiation (λ =
Single spot on TLC (silica gel plate,
1.5418 Å) at a scanning rate of 0.3°/min. EtOAc-MeOH 1:1, Rf = 0.2).
Oxygen permeability measurements were
done on sheet silicate loaded polyethyl- 2.4. Benzyl-[3-[2-(2,6-diethyl-4-
ene (PE) films at the Frauenhofer Institute hydroxy-2,3,6-trimethylpiperidin-1-
for Chemical Engineering and Packaging yloxy)-2-methylpropionyl-amino]-
(D-85354 Freising) according to DIN 53 propyl]-dimethylammonium
380 (T3). The PE pellets (Lupolen 2420
from BASF) were extruded with 3–5% ex- Benzylchloride (0.87g, 6.87 mmol) was Synthetic Layered Silicate ‘Optigel
foliated sheet silicates at 180 °C and the added to a solution of 8 (2.2 g, 5.7 mmol) in SH’
melt blown at 200 °C to an approx. 30 µm acetonitrile (3 ml). The mixture was stirred 60.25 g (117.4 mmol, 1.2 equiv. relative
thick film. The oxygen diffusing through for 19 h at room temperature and then evap- to the ion exchange capacity of the sheet
this film was measured with a chemical orated. The residue was treated with dieth- silicate) 9 was dissolved in a mixture of 850
ylether to remove the excess of benzylchlo- ml distilled water and 150 ml EtOH. 125.4
chloride (9)
2.7. Intercalation of 9 into the
sensor.
[35]
ride, the solid was filtered off and dried to g Optigel SH (containing 15 wt.% water
)
2.2. 2-Bromo-N-
afford 2.9 g (99%) of the title compound as was dispersed in this solution and stirred
(3-dimethylaminopropyl)-2-
methylpropionamide (7)
a colorless amorphous solid.
for 24 h at r.t. The dispersion was filtered,
MS (Infusion-ESI) for the cation washed with water and dried in vacuo (3 d,
2-Bromoisobutyryl bromide (23.0 g, 0.1 C28H50N3O3 (476.73): found M+ = 476.4. 50 °C, 0.01 mbar).Yield: 154.2 g of a white
mol) was added dropwise over 50 min to 1H-NMR (300 MHz, CDCl3), mixture of powder.
a solution of 3-dimethylaminopropylamine diastereoisomers: 7.70–7.55 (m, 2 ArH),
Weight loss determined by TGA (50–
(25.2 ml, 0.2 mol) in THF (50 ml), while
7.50–7.30 (m, 3 ArH), 7.05–6.85 (bs, NH), 600 °C): 31.5 wt%, corresponding to 66.1
keeping the temperature between 0–10 °C. 4.91 (bs, CH2Ph), 4.30–4.10 (m, CH(OH), mmol 9 in 100 g intercalated synthetic sheet
The mixture was stirred for another 3 h at
3.70–3.50 (m, CH2), 3.34 (s, N+(CH3)2), silicate.
room temperature and the THF was then
2.50–0.60 (m, 33H).
evaporated in vacuo. Water (20 ml) was
2.8. Exfoliation of Natural Sheet
Silicate Nanofil EXM 588 Containing
the Intercalated Cationic
added to the residue and the mixture was 2.5. Bulk Polymerization of
extracted with t-butyl-methyl ether (2 × 30 n-Butylacrylate with the
ml) and ethylacetate (30 ml). The combined
Alkoxyamines 8 and 9
A 50 ml three-necked round bottom Acrylate (Scheme 3)
solution (10 ml), dried over MgSO4 and flask with N2 in- and outlet, reflux condens- In a typical example, 57.8 g Nanofil
evaporated to afford 24.1 g (96%) of the er and magnetic stirring was charged with EXM 588 intercalated with alkoxyamine 9
Alkoxyamine 9 via NMP of n-Butyl
extracts were washed with saturated NaCl
title compound as colorless oil.
10 g (78 mmol) n-butyl acrylate, 0.451 g (32.9 mmol) was dispersed in a mixture of
1H-NMR (300 MHz, CDCl3): 8.51 (bs, (1.17 mmol, 1.5 mol%) of the alkoxyamine 308 g (2.4 mol) n-butyl acrylate and 924 g
NH), 3.39–3.34 (m, CH2), 2.45 (t, J = 6 Hz, 8 resp. 0.599 g (1.17 mmol, 1.5 mol%) of 9 2-methoxypropyl acetate and homogenized
CH2), 2.25 (s, 2 × CH3), 1.94 (s, 2 × CH3), and 2 g N,N-dimethyl-acetamide. The solu-
with an ultraturrax mixer and ultrasonic
1.72–1.64 (m, CH2).
tion was deoxygenated by five evacuation/ bath. The dispersion was transferred to a

POLYMERS AND COLLOIDS DIVISION
802
CHIMIA 2008, 62, No. 10
weight of the polymer attached to the in-
dividual sheets, a sample (150 mg) of this
solid was refluxed with 15 ml 0.1 M LiBr
O
Na+
Na+
Na+
Na+
Intercalation
Na+
Na+
with 9
solution in THF during 17 h. After filtra-
tion, the molecular weight was determined
by GPC in THF: Mn = 3600, Mw = 5280,
PD = 1.47.
N
H
O
N
OH
EtOH / H2O
24 h, RT
Ph
N+
2.9. Exfoliation of Synthetic Sheet
Montmorillonite
Montmorillonite with intercalated 9
Silicate Optigel SH Containing the
Intercalated Cationic Alkoxyamine 9
via NMP of n-Butyl Acrylate.
62.3 g Optigel SH intercalated with
O
alkoxyamine 9 (41.2 mmol) was dispersed
in a mixture of 386.8 g (3.02 mol) n-butyl
acrylate and 1160 g 2-methoxypropyl ac-
n-BuO
O
n-Butyl acrylate
140 °C
O
N
OH
n
N
H
etate and homogenized with an ultraturrax
Exfoliation
Ph
mixer and ultrasonic bath. The dispersion
N+
was transferred to a 2.5 l three-necked
round bottom flask with reflux condenser,
magnetic stirring and nitrogen in- and out-
let. After evacuation and purging five times
with N2, polymerization was performed
under N2 for 20 h at 140 °C under vigor-
ous stirring. After that time, the monomer
Exfoliated montmorillonite nanocomposite
Scheme 3. Intercalation of 9 and exfoliation through nitroxide-mediated polymerization
1
conversion, determined by H-NMR, was
25%. The solvents were evaporated in the
rotavap and the residue extracted in a Sox-
hlet with 2 l EtOAc during 18 h to remove
[37]
all non-attached poly(n-butyl acrylate).
After drying in vacuo, 71.4 g white solid
was obtained.
Weight loss in TGA (50–600 °C): 40.8
wt%. X-ray analysis: Only peaks at 2Θ
>10° are found, arising from atomic dis-
tances within the layers and not between
the layers, indicating complete exfoliation.
The polymer was cleaved off the sheet sili-
cate by refluxing a sample (150 mg) in 15
ml 0.1 M LiBr solution in THF during 17
h. After filtration, the molecular weight de-
termined by GPC in THF, was: Mn = 1210,
Mw = 2300, PD = 1.90.
3. Results and Discussion
3.1. Synthesis of the Cationic NMP
Alkoxyamine Initiator 9
The target alkoxyamine 9 (Scheme 2)
was obtained via straightforward quater-
nization of the key intermediate 8 with
Fig. 2. X-ray diffractograms of a) hydrated Na+-montmorillonite (Nanofil EXM588), b)
intercalated montmorillonite with 9, and c) exfoliated montmorillonite made by NMP
of n-butyl acrylate with intercalated 9 (sample A in the Table)
benzyl chloride in acetonitrile. The mate-
rial was obtained as a colorless amorphous
solid after the excess of benzyl chloride was
washed away with diethyl ether. It is mod-
erately soluble in water or aqueous ethanol.
No further purification was attempted. In
fact, chromatography would be difficult
given the salt character and the compound
2.5 l three-necked round bottom flask with ing 18 h to remove all non-attached poly(n-
[36]
reflux condenser, magnetic stirring and ni- butyl acrylate). After drying in vacuo,
trogen in- and outlet. After evacuation and 136 g resin-like material was obtained.
failed to crystallize, presumably because
Weight loss in TGA (50–600 °C): 64.4
purging five times with N2, polymeriza-
it is present in form of several diastereo-
isomers (see refs [27,38] for the discussion
of the isomeric composition of 2,6-diethyl-
tion was performed under N2 for 20 h at wt%. X-ray analysis (Fig. 2c): Only peaks
140 °C under vigorous stirring. After that at 2Θ >10° are found, arising from atomic
time, the monomer conversion, determined distances within the layers and not between
1
2,3,6-trimethylpiperidine nitroxides). The
intermediate 8 was prepared via the atom
transfer radical addition (ATRA) coupling
by H-NMR, was 30%. The solvents were
the layers, indicating complete exfoliation.
evaporated in the rotavap and the residue In order to cleave off the polymer from the
extracted in a Soxhlet with 2 l EtOAc dur- sheet silicate and to analyze the molecular

POLYMERS AND COLLOIDS DIVISION
803
CHIMIA 2008, 62, No. 10
[39]
[28]
Table. Exfoliation of natural (Nanofil EXM 588) and a synthetic (Optigel SH) sheet silicates, intercalated
with 9, using NMP of n-butyl acrylate (BA) in methoxypropylacetate (MPA) solution at 140 °C
reaction
of the known
nitroxide 5
with the activated bromoamide 7 promot-
ed by excess amount of Cu(i) ions in the
presence of N,N,N’,N’’,N’’-pentamethyl-
Sample Sheet
No. silicate
Mol.% 9 in
BA, wt % BA Conversion
Time,
Mn (calc) Mn, Mw
PD (exp)
wt.% Polymer
in sheet
diethylenetriamine (PMDTA) ligand. The
in MPA
silicate
bromoamide 7 was readily obtained via
acylation of 3-dimethylamino-propylamine
with 2-bromoisobutyryl bromide.
A
B
C
D
Nanofil
EXM 588
1.37, 25%
20 h, 30%
24 h, 30%
24 h, 33%
20 h, 25%
3280
6240
11040
2660
3600, 5280
1.47
64
Nanofil
EXM 588
0.67, 48%
0.40, 100%
1.47, 25%
6360, 12220
1.92
78
89
41
3.2. Polymerization of n-Butyl
Acrylate with Alkoxyamines 8 and 9
Both alkoxyamines 8 and 9 are efficient
mediators of bulk NMP of n-butyl acrylate
affording 71 resp. 65 % monomer conver-
sion after 7 h polymerization at 140 °C.
Nanofil
EXM 588
13700, 28700
2.09
Optigel
SH
1210, 2300
1.90
The calculated Mn-values correspond well
to those obtained and the polydispersities
(PD = 1.21 resp. 1.34) of the polymers are
rather low.
ionic alkoxyamine 9 can be calculated to (Mn, Mw) and polydispersity PD were de-
be 0.57 mmol/g intercalated natural sheet termined by GPC in THF (relative to PS-
silicate Nanofil and 0.66 mmol/g interca- standards).
3.3. Intercalation of Natural and
Synthetic Sheet Silicate with
Cationic Alkoxyamine 9
lated synthetic sheet silicate Optigel. These
The results of all exfoliation experiments
values are slightly lower than the specified are summarized in the Table. Depending on
Cation exchange of the natural and syn- values of 0.61 and 0.78 mmol/g, respec- the monomer to initiator ratio and the mono-
thetic sheet silicates and intercalation of the tively. Presumably, the quite large cation mer concentration, the number average mo-
cationic alkoxyamine 9 was conducted in of 9 cannot replace all cations of the na- lecular weight Mn of the poly(n-butyl acry-
aqueous ethanol (85% water).Almost quan- tive sheet silicates completely due to steric late), attached to the individual sheet silicate
titative exchange of Na+ ions with the cat- constraints.
ionic alkoxyamine is proven by gravimet-
ric analysis and powder X-ray diffraction. 3.4. Exfoliation of Natural and
layers, could be adjusted to be between 1000
and 14000 g/mol, translating into a polymer
weight content of 41–89%. Due to steric
The unmodified montmorillonite (Fig. 2a)
Synthetic Sheet Silicates by NMP
constraints between the layers, which is par-
shows the most important reflection peak at
5–20 wt.% of two intercalated sheet ticularly the case at the beginning of the po-
2Θ = 7.1°, corresponding, as calculated ac- silicates, containing the immobilized 9, lymerization, the monomer conversion was
cording to Bragg’s equation (nλ= 2.d.sinΘ), were well dispersed in a mixture of n-butyl lower than in the corresponding bulk po-
to the original distance between sheets of d acrylate (monomer) and 2-methoxypropyl lymerization experiments (see Experimental
= 1.24 nm. Fig. 2b displays the powder X- acetate(solvent)andpolymerizedwithvig- Section) and larger PDs were obtained. Nev-
ray of the montmorillonite sample, interca- orous stirring under N2. In order to avoid ertheless, the polymerization was somewhat
lated with 9. The maximum reflection peak
a highly viscous reaction mixture or even controlled since the calculated molecular
is shifted to 2Θ = 3.86°, corresponding to gelling, the appropriate amount of solvent weights are in quite good agreement with
the new interlayer distance of d = 2.21 nm. was chosen and the monomer conversion the obtained values.
Note that the peak at 2Θ = 7.1° has com- was kept below 33%. After exfoliation,
An exception is the exfoliation of the
pletely disappeared whereas the peaks at the dispersion was very homogeneous and synthetic silicate Optigel SH where most of
higher 2Θ-values are still there since they stable since rigorous centrifugation gave the formed polymer is not attached and the
represent atomic distances within the lay- only a very small amount (<8%) of sedi- molecular weight of the attached polymer
ers. Therefore, compared with the native
mented product, probably non-exfoliated approximately two times lower than calcu-
sheet silicate, an increase of the interlayer sheet silicate. Evaporation of all monomer/ lated. In synthetic sheet silicates, the area of
distance of 0.97 nm is obtained, corre- solvent in high vacuo led to waxy or solid the individual silicate layers are generally
sponding approximately to the size of the residues. These were extracted with ethyl
much smaller (<100 nm in x, y-direction
intercalated 9. Unfortunately, for unknown acetate to remove all non-attached poly- compared to 200–500 nm for montmorillo-
[40]
reasons, we were not able to measure a mer and dried. TGA analysis revealed
nite) and therefore we assume that most of
meaningful X-ray diffraction diagram for the polymer layered silicate composition the exfoliated product is lost during soxhlet
the synthetic sheet silicate Optigel SH in- and X-ray analysis gave only peaks at 2Θ extraction, leaving just sheet silicate with
tercalated with 9.
>10°, arising from atomic distances with- short poly(n-butyl acrylate) chains behind.
The exact amount of intercalated 9 was
determined by thermogravimetric analysis Since the smallest 2Θ-value measurable initiator,
α,α'-azodiisobutyramidine-
(TGA) on completely dried samples. Upon
with our instrument is 3°, the d spacing dihydrochloride was intercalated into
in the layers and not between the layers.
Finally, a classical thermal radical
heating from r.t. to 600 °C (heating rate: 10 is >3 nm, indicating complete exfoliation. Nanofil EXM 588 in order to directly com-
°C/min), a weight loss of 27.1% was found In order to analyze molecular weight and pare the above-described exfoliation by
for the intercalated Nanofil and 31.5% for polydispersity of the immobilized poly- NMP with exfoliation via uncontrolled po-
[41]
the intercalated Optigel, corresponding to mer, it was cleaved off the silicate surface lymerization. The TGA and GPC results
the amount of organic material between the by refluxing the nanocomposite suspen- showed clearly that only a small fraction
layers. These values agree well with the
sion in THF in the presence of LiBr for of the intercalated thermal radical initiator
[15]
difference in solution concentrations mea- an extended period of time. This caused
is active in polymerization, which is strik-
sured by UV-spectroscopy before and after a cation exchange as the Li+-ions replace ingly different to the exfoliation by NMP.
the intercalation process and the final the polymer’s cationic end groups, thus
[34]
Therefore, it can be concluded, that the
yields. From these data, the ion-exchange separating the polymer from the insoluble exfoliation of montmorillonite-type layered
capacities of the sheet silicates for the cat- silicate. After filtration molecular weight silicates with efficient initiation in all layers

POLYMERS AND COLLOIDS DIVISION
804
CHIMIA 2008, 62, No. 10
[1] E. P. Giannelis, Adv. Mater. 1996, 8, 29.
[2] C. Zilg, R. Thomann, R. Muelhaupt, J.
[30] P. Nesvadba, Chimia 2006, 60, 832.
[31] C. Auschra, E. Eckstein, R. Knischka, P.
due to intercalated 9 followed by controlled
radical polymerization is a much better
method to obtain real nanocomposites in
Finter, Adv. Mater. 1999, 11, 49.
Nesvadba, Asia Pac. Coat. J. 2003, 16,
20.
[3] C. J. G. Plummer, L. Garamszegi, Y.
high yield. The amount of non-exfoliated
Leterrier, M. Rodlert, J.-A. E. Maanson, [32] A. Muehlebach, P. Nesvadba, A.
layered silicate is small and the formed
polymer has a much lower, controlled mo-
Chem. Mater. 2002, 14, 486.
Kramer, WO Patent Publication No.
WO2004000809, 2003.
[4] C. Zilg, F. Dietsche, B. Hoffmann, C.
lecular weight. This translates into a higher
Dietrich, R. Mulhaupt, Macromol. Symp. [33] Measured by the weight loss between 50
sheet silicate content of the nanocomposite
and a lower dispersion viscosity, which is
very important for most applications. Fur-
thermore, the exfoliated sheet silicates can
2001, 169, 65.
and 200 °C in TGA. The cation exchange
[5] D. Schmidt, D. Shah, E. P. Giannelis,
Curr. Opin. Solid State Mater. Sci. 2002,
6, 205.
capacity is 0.61 mmol/g or 97.4 mmol
exchangeable monovalent cations in the
whole sample.
[6] Q. T. Nguyen, D. G. Baird, Adv. Polym.
[34] Samples of the dispersion were taken in
be re-dispersed easily in solvents like THF,
Technol. 2006, 25, 270.
regulartimeintervalsandtheconcentration
of the alkoxyamine 9 in the supernatant
clear solution (after centrifugation) was
EtOAc, BuOAc, toluene etc. and the dis-
persions can be centrifuged during many
hours without any sedimentation. Since
even diluted dispersions (e.g. 1% in EtOAc)
remain stable for years without visible sedi-
mentation, we conclude that near complete
exfoliation to nm size particles took place.
[7] M. Okamoto, Int. Polym. Process. 2006,
21, 487.
[8] P. A. Wheeler, J. Wang, J. Baker, L. J.
Mathias, Chem. Mater. 2005, 17, 3012.
[9] H. Van Olphen, ‘An Introduction to Clay
Colloid Chemistry’, Willey-Interscience,
New York, 1977.
determined by UV-spectroscopy at λ =
245 nm, showing that the concentration
of the remaining (non-intercalated) 9
is constant (and therefore intercalation
terminated) after approx. 20 h at these
[10] P. B. Messersmith, E. P. Giannelis, Chem.
Mater. 1994, 6, 1719.
[11] H. Boettcher, M. L. Hallensleben, S. Nuss,
H. Wurm, J. Bauer, P. Behrens, J. Mater.
Chem. 2002, 12, 1351.
conditions.
[35] Measured by the weight loss between 50
and 200°C in TGA. The cation exchange
4. Application Results: Reduction of
Oxygen Permeability in PE foils
capacity is 0.78 mmol/g or 97.8 mmol
exchangeable monovalent cations in the
whole sample.
[12] D. Kubies, N. Pantoustier, P. Dubois, A.
Polyethylene foils were extruded with 3
wt.% of the unmodified Nanofil EXM 588
and 5 wt% of the exfoliated montmorillo-
Rulmont, R. Jerome, Macromolecules [36] The total amount of extractables was 7.1
2002, 35, 3318.
[13] C. J. B. Mott, Catal. Today 1988, 2, 199.
g which corresponds to ca. 8% of the
polymer formed.
nite-type sheet silicate (sample A, Table).
The incorporation of 3% unmodified sheet
[14] L. P. Meier, R. A. Shelden, W. R. Caseri, [37] The total amount of extractables was 68.1
U. W. Suter, Macromolecules 1994, 27,
g which corresponds to ca. 70% of the
silicate increased the oxygen permeability
around 10% (probably due to voids at the or-
ganic/inorganic interphase because of lack
of compatibility), whereas 5% A decreased
the O2-permeability to about 75–80% of the
original value. Since the exfoliated plate-
lets incorporated into the PE matrix are pre-
sumably not ordered the effect is relatively
small and thus probably not sufficient for
commercial applications.
1637.
polymer formed.
[15] U. Velten, R. A. Shelden, W. R. Caseri, U. [38] H. Fischer, A. Kramer, S. R. A. Marque,
W. Suter,Y. Li, Macromolecules 1999, 32,
3590.
P. Nesvadba, Macromolecules 2005, 38,
9974.
[16] M. Albrecht, S. Ehrler, A. Muhlebach, [39] J.-L. Couturier, O. Guerret, T.
Macromol. Rapid Commun. 2003, 24,
382.
Senninger, WO Patent Publication No.
WO2000061544, 2000.
[17] W. A. Braunecker, K. Matyjaszewski, [40] The relative amount of extracted non-
Prog. Polym. Sci. 2007, 32, 93.
[18] D. H. Solomon, J. Polym. Sci., Part A:
Polym. Chem. 2005, 43, 5748.
attached poly(n-butyl acrylate) was small
in the exfoliated montmorillonite Nanofil
(ca. 8%) but large (ca. 70%) in the
[19] C. J. Hawker, A. W. Bosman, E. Harth,
Chem. Rev. 2001, 101, 3661.
[20] C. Konn, F. Morel, E. Beyou, P. Chaumont,
E. Bourgeat-Lami, Macromolecules 2007,
40, 7464.
synthetic sheet silicate.
[41] α,α'-Azodiisobutyramidine-dihydro-
chloridewasintercalatedintoNanofilEXM
5. Conclusions
588 to obtain an initiator concentration of
99.2 mg per 1 g of intercalated product
We have disclosed an efficient route
towards nanocomposite based on the inter-
calation of a natural and a synthetic sheet
silicate by cation exchange with cationic
alkoxyamine initiator/regulator 9. Other
suitable cationic alkoxyamines are accessi-
ble in just a few synthetic steps. Nanocom-
posites were obtained by thermally induced
NMP of n-butyl acrylate, leading to almost
complete exfoliation of the silicate layers.
Such nanocomposites show increased gas
barrier properties. Screening for other pos-
sible applications of these unique hybrid
nanomaterials is currently in progress.
[21] J. Pyun, K. Matyjaszewski, Chem. Mater.
2001, 13, 3436.
[22] M. W. Weimer, H. Chen, E. P. Giannelis,
D.Y. Sogah, J. Am. Chem. Soc. 1999, 121,
1615.
(0.49 mmol/g). n-Butyl acrylate (9.5 g)
was then polymerized under N2 with 0.5
g of this intercalated sheet silicate for 3 h
at 80 °C. The dispersion was diluted with
toluene (240 ml) and centrifuged to give
0.35 g (70%) of sedimented intercalated
Nanofil EXM 588, as verified by TGA and
X-ray analysis. The remaining monomer
and solvent in the supernatant was
evaporated in vacuo. The residue (2.00
g) contained, according to TGA-analysis,
93% polymer and only 7% sheet silicate.
[23] J. Di, D. Y. Sogah, Macromolecules 2006,
39, 5052.
[24] D. Benoit, V. Chaplinski, R. Braslau, C.
J. Hawker, J. Am. Chem. Soc. 1999, 121,
3904.
[25] C. Le Mercier, S. Acerbis, D. Bertin,
F. Chauvin, D. Gigmes, O. Guerret, M.
Lansalot, S. Marque, F. Le Moigne, H.
GPC analysis of this residue showed ca.
50% ‘free’ polymer (not attached to the
sheet silicate) with Mn = 790 000 (PDI
Fischer, P. Tordo, Macromol. Symp. 2002,
182, 225.
[26] P. Nesvadba, A. Kramer, M.-O. Zink, DE
Patent Publication No. DE99-19949352,
2000.
= 1.90) whereas the attached polymer
(ca. 50%, detachment by reflux with 0.1
Acknowledgements
We thank Ms. F. Modoux for the X-ray
measurements and Mr. C. Gerst for providing the
M LiBr in THF) showed Mn = 660 000
(PDI = 2.06), indicating uncontrolled
polymerization.
[27] P. Nesvadba, L. Bugnon, R. Sift, J. Polym.
TGA and DSC data. Dr. Engelhardt from Süd-
Sci., Part A: Polym. Chem. 2004, 42,
Chemie, Germany is gratefully acknowledged
for the Nanofil EXM 588 and Optigel SH sheet
silicate samples.
3332.
[28] A. Kramer, P. Nesvadba, DE Patent
Publication No. DE99-19909767, 1999.
[29] W. Rutsch, M. A. Cech, Chimia 2007, 61,
33.
Received: March 27, 2008