Novel fluorinated block copolymer architectures fuelled by atom transfer radical polymerization

Article abstract of DOI:10.1016/j.jfluchem.2004.11.002

Block copolymers based on poly(pentafluorostyrene), PFS, in various numbers and of different lengths, and polystyrene are prepared by atom transfer radical polymerization (ATRP). Di- and triblock copolymers with varying amounts of PFS were synthesized employing either 1-phenylethylbromide or 1,4-dibromoxylene as initiators for ATRP. Diverse bromo(ester) (macro)initiators were also devised and involved in the formulation of fluorinated pentablock as well as amphiphilic triblock copolymers with a central polyether segment. Amphiphilic star-shaped fluoropolymers, hydrophobic fluorinated nanoparticles, or segmented fluorinated star-shaped block copolymers are further designed by use of different multifunctional initiators. The composition of the novel materials with PFS is determined by combination of SEC and ¹H NMR. Glass transition temperatures and thermal stabilities of the hydrophobic star-shaped PFSs on a six arm dipentaerythritol core are investigated in a wide range of molecular masses and further discussed.

Full text of DOI:10.1016/j.jfluchem.2004.11.002

Journal of Fluorine Chemistry 126 (2005) 241–250
www.elsevier.com/locate/fluor
Novel fluorinated block copolymer architectures fuelled by
atom transfer radical polymerization
*
Katja Jankova, Søren Hvilsted
Danish Polymer Centre, Department of Chemical Engineering, Technical University of Denmark,
Building 423, DK-2800 Kgs. Lyngby, Denmark
Received 2 June 2004; received in revised form 19 October 2004; accepted 8 November 2004
Available online 10 December 2004
Abstract
Block copolymers based on poly(pentafluorostyrene), PFS, in various numbers and of different lengths, and polystyrene are prepared by
atom transfer radical polymerization (ATRP). Di- and triblock copolymers with varying amounts of PFS were synthesized employing either 1-
phenylethylbromide or 1,4-dibromoxylene as initiators for ATRP. Diverse bromo(ester) (macro)initiators were also devised and involved in
the formulation of fluorinated pentablock as well as amphiphilic triblock copolymers with a central polyether segment. Amphiphilic star-
shaped fluoropolymers, hydrophobic fluorinated nanoparticles, or segmented fluorinated star-shaped block copolymers are further designed
by use of different multifunctional initiators. The composition of the novel materials with PFS is determined by combination of SEC and 1H
NMR. Glass transition temperatures and thermal stabilities of the hydrophobic star-shaped PFSs on a six arm dipentaerythritol core are
investigated in a wide range of molecular masses and further discussed.
# 2004 Elsevier B.V. All rights reserved.
Keywords: Pentafluorostyrene; ATRP; Bromoester macroinitiators; Block copolymers; Amphiphilic; Contact angle; Polyether blocks; Dipentaerythritol;
Star-shaped polymers; Fluorinated nanoparticles
1. Introduction
the form of block copolymers for generation of low-energy
surfaces [9–19] or as surface-modified membranes [20].
Most fluorinated block copolymers with well-defined
structures applied for generation of low-energy surfaces are
either polystyrene based and initially prepared by living
anionic polymerization [11,12,14–16,18,19] or poly(butyl
methacrylate-co-perfluoroalkyl acrylate) [17] synthesized
by atom transfer radical polymerization (ATRP). In addition,
block copolymers with poly(ethylene oxide) and poly(per-
fluoroalkyl methacrylate) [21–23] have been prepared by
ATRP. 2-Perfluoroalkyl containing initiators were likewise
employed in ATRP of copolymers [24,25]. Low critical
surface tensions have also been determined in fluorinated
poly(amide urethane) block copolymers with fluorinated
side chains prepared from diacid chlorides [9,10]. In the
styrene based polymers the fluorination was obtained by
functionalization of polystyryllithium and subsequent
Williamson reactions [14,15], by deprotection of poly(4-
tert-butyldimethylsiloxyloxystyrene) followed by William-
Fluorinated polymers have always attracted significant
attention due to high thermal stability, excellent chemical
resistance, superior oil and water repellence or low
flammability in addition to low refractive indexes [1].
Additionally, thin films of fluorinated polymers have
recently been researched for low optical loss to be exploited
in wave-guiding devices [2–5]. Also, low permittivity or low
dielectric constants [6] are desirable properties for the optics
and electronics industries. Furthermore, amphiphilic fluor-
ine containing polymers appear to have a considerable
potential as electrolyte materials for Li+ conductivity in
solid-state lithium-ion polymer battery applications [7,8].
Other recent, notable applications of fluoropolymers are in
* Corresponding author. Tel.: +45 45252965; fax: +45 45882161.
E-mail address: sh@kt.dtu.dk (S. Hvilsted).
0022-1139/$ – see front matter # 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2004.11.002

242
K. Jankova, S. Hvilsted / Journal of Fluorine Chemistry 126 (2005) 241–250
son reactions [15], or via oxidative hydroboration of
isoprene blocks succeeded by esterification with perfluori-
nated acid chloride [11,12,14,19]. Finally, examples of
TEMPO-mediated controlled radical polymerization of
fluorinated alkoxymethylstyrene homo- [19] or block
copolymers [26] also exist.
hydroxyl compounds leads to the syntheses of novel
fluorinated amphiphilic star-shaped materials that can be
regarded as fluorinated nanoparticles. Finally, we elude to
some possible applications of these novel materials that are
currently under investigation.
Most recently ATRP has proved to be the most rapidly
growing area of polymer chemistry with numerous
possibilities for polymerization of functional monomers in
a controlled fashion, extended use of initiators with
functional groups, and the potential polymer end group
transformation [27]. Furthermore, ATRP has appeared as a
valuable tool with great design flexibility awarding often full
control of complex polymer composition, topology, and
functionality including viable methods for preparation of
molecular composites [28]. Thus, a number of monosub-
stituted styrenes, e.g. 4-fluoro- and 4-trimethylfluorostyrene,
have been polymerized by ATRP [29]. That prompted us to
investigate the polymerizability of the fully phenyl
fluorinated styrene, 2,3,4,5,6-pentafluorostyrene (FS), by
ATRP [30]; another brief report on FS polymerization by
ATRP also exists [31]. The rewarding and fast poly-
merization of FS has been successfully followed by ATRP
of 2,3,5,6-tetrafluoro-4-methoxystyrene (TFMS) [32,33].
Most recently, newly synthesized highly fluorinated
fluoroalkoxy styrene monomers, 2,3,5,6-tetrafluoro-4-(2,2,
3,3,3-pentafluoropropoxy)-styrene and 2,3,5,6-tetrafluoro-
4-(2,2,3,3,4,4,5,5,6,6,7,7,8,8,8-pentadecafluoroctaoxy)styr-
ene were found to produce polymers capable of forming low
surface energy materials [34].
2. Results and discussion
2.1. Block copolymers with varying fluorine content
The controlled characteristics of PFS and PTFMS were
exploited in the preparation of a number of block
copolymers [33]. The conservation of the bromine reactivity
in the isolated homopolymers was indirectly proved through
the efficient ability to act as macroinitiators (MIs). Fig. 1a
illustrates a PFS-b-PS (2) block copolymer where the two
end groups, 1-phenylethyl and bromine, originating from the
initiator, 1-phenylethyl bromide, initially employed when
the PFS macroinitiator was prepared, are shown. The block
copolymers suffered only slight increases in the total
polydispersity index (PDI = Mw/Mn) as compared with those
of the initiating blocks. Furthermore, the applicability of the
MIs in combination with the high reactivity of the monomers
in this way provides a flexible tool for design of novel block
copolymers. Moreover, the sequence of the blocks can easily
be varied; in fact, among the mixtures of FS, TFMS and
styrene (st.) all six possible combinations of block
copolymers have been prepared. In practice, the decision
on the block sequence adopted often depends on the
solubility of the particular MI, which in turn also depends on
the block length.
These findings open up new design avenues for
preparation of novel fluorinated polymer materials. We
here demonstrate how the potential of ATRP has been
further exploited for synthesis of block copolymers (both
hydrophobic and amphiphilic) with varying fluorine content.
The application of multifunctional initiators based on
BAB type block copolymers have previously been
synthesized in our group by ATRP using 1,4-dibromoxylene
as the difunctional initiator [35]. The triblock copolymer in
Fig. 1b, PFS-b-PS-b-PFS (4), was produced by two sequential
Fig. 1. Examples of diblock, triblock, and pentablock fluorocopolymers.

K. Jankova, S. Hvilsted / Journal of Fluorine Chemistry 126 (2005) 241–250
243
Table 1
Composition, molecular weights and contact angles of di-, tri-, and pentablock fluoropolymers
Polymer
Mn (RI)a
Mw/Mn (RI)a
Mn (LS)a
Mw/Mn (LS)a
PFS by 1H NMR (wt.%)
Contact angleb (8)
PFS-Br (1)
PFS-b-PS (2)
Br-PS-Br (3)
PFS-b-PS-b-PFS (4)
PS-b-F16-b-PS (5-1)
PS-b-F16-b-PS (5-2)
PFS-b-PS-b-F16-b-PS-b-PFS (6-1)
PFS-b-PS-b-F16-b-PS-b-PFS (6-2)
a
10900
21800
17400
25300
6600
1.15
1.24
1.14
1.37
1.17
1.17
1.38
1.36
108c
107
95c
55
19
21400
32200
1.11
1.34
102
100
11100
19200
23900
56
58
RI: refractive index detector; LS: light scattering detector.
Advancing contact angle of water droplet.
Ref. [34].
b
c
ATRPs where the dibromo ends in the isolated Br-PS-Br (3)
served as initiator for the subsequent polymerization with FS
furnishing the triblock copolymer (4). Based on the PS
calibration and the number average molecular weight (Mn) of
the MI, the block copolymercomposition listedin Table 1 was
calculated by 1H NMR as previously performed for diblock
copolymers [30]. From these data the Mn of the block
copolymer (4) was calculated by 1H NMR to be 21,500. The
examples presented in Fig. 1c, PFS-b-PS-b-F16-b-PS-b-PFS
(6-1 and 6-2), are formally pentablock copolymers of the
CBABC type, although the A block (F16 represents the
fluorinated decanediol moiety) is very short. Nevertheless, in
this manner it is possible to have three fluorinated blocks
separated by PS blocks in the same copolymer. This type of
block copolymer was also synthesized by two individual
ATRP initially with St and then with FS. However, in this case
a suitable initiator was first produced by converting a highly
fluorinated decanediol, namely the 1H,1H0,10H,10H0-per-
fluorodecane-1,10-diol to the corresponding bromoester (A-
block) as described in the following section, and outlined in
Scheme 1 in case of the dihydroxypolyethers. Hereafter the
procedure resembled the normal procedure for block
copolymer synthesis. Finally, it should be noted that the
fluorine content of block copolymers can easily be varied
when PS constitutes one block.
Since the polymerizations perform in a controlled fashion,
the fluorine content can to a certain extent be controlled
through the original fluorine monomer to initiator ratio.
Continuous investigations of the fluorinated materials’
surface properties were conducted by determination of the
advancing contact angles (CA) of water drops on spin coated
surfaces. The results of the CA determination listed in Table 1
show that the fluorine block introduced imparts higher
hydrophobicity of the material surface as seen from the
increase of the CA of 2 and 5 as compared with the non-
fluorinated 3. Even the small content of the central segment of
16 aliphatic fluorine atoms in 5-2 corresponding to less than
4 wt.% ofthepolymer increasestheCAof PSfrom 95to1008.
The CA of the block copolymer 2 with 55 wt.% PFS reaches
almost the CA of a homopolymer of FS (1088 [34]). Thus it
seems that the fluorinated blocks preferentially segregate at
the air/film interface when spin coated, as confirmed recently
for other fluorinated block copolymers [34].
2.2. Fluoropolymers with polyether blocks
A strategy for preparation of BAB triblock copolymers
with a poly(ethylene glycol) (PEG) A, middle block flanked
by PS end blocks (B) by ATRP was introduced already in
1998 by us [36,37]. The concept is based on quantitative
Scheme 1. Preparation of amphiphilic triblock fluorocopolymers by ATRP.

244
K. Jankova, S. Hvilsted / Journal of Fluorine Chemistry 126 (2005) 241–250
1
Table 2
combination of SEC and H NMR. In addition, the SEC
revealed PDIs < 1.22 for the triblock copolymers.
Composition and molecular weights of amphiphilic triblock copolymers of
PFS with a central polyether block
These amphiphilic triblock copolymers have very
interesting material properties. For example PFS-b-
PEGPG-b-PFS (19) is a good film forming material,
probably due to a strong phase separation in which the
PFS domains act as physical crosslinks. In fact, even a much
lower PFS content on the order of 3–9 wt.% ensures the
formation of an elastomeric material [38]. We believe that
the molecular construction is quite similar to the thermo-
plastic elastomers where two PS outer blocks flank an
elastomer central block. However, in this case an even
stronger phase separation between the fluoropolymer blocks
and the polyether segments is envisioned due to the strong
amphiphilic character of the material. Furthermore, the
polyether block in this material has demonstrated good Li+
complexation when mixed with Li salts while the composite
still maintains the favourable film forming properties.
Moreover, this composite has demonstrated good Li+
conductivity (10À4.9 to 10À5.1 S cmÀ1 at 20 8C) over a
viable temperature range (À40 to +100 8C) [38]. Addition-
ally, it seems that in some instances the fluorinated blocks
ensure good contact between the electrode materials and the
electrolyte polymer. Consequently they appear as potential
candidates for electrolyte materials in solid-state Li+
polymer battery applications.
a
a
Polymer
Mn,SEC
Mw/Mn
PFS (wt.%) Mn,NMR
PEG2 (7)
Br-PEG2-Br (8)
PFS-b-PEG2-b-PFS (9)
PFS-b-PEG2-b-PFS (10)
1900
2400
5300
6200
1.02
1.02
1.05
1.05
80
84
9800
12200
PEG5 (11)
Br-PEG5-Br (12)
PFS-b-PEG5-b-PFS (13)
4500
4800
7600
1.01
1.02
1.20
51
39
39
9500
17200
15800
PEG10 (14)
Br-PEG10-Br (15)
PFS-b-PEG10-b-PFS (16)
10300
11000
24300
1.12
1.16
1.20
PEGPG (17)
Br-PEGPG-Br (18)
PFS-b-PEGPG-b-PFS (19) 15300
9400
9600
1.02
1.05
1.22
a
By SEC in THF employing PEG calibration.
conversion of the terminal hydroxyl groups in PEG to 2-
bromo propionates. The bromine in this environment serves
as an effective initiator for ATRP. The resulting bromine
ester terminated PEG then performs as a difunctional MI
employed in the preparation of PS-b-PEG-b-PS copolymers.
This strategy was adapted and exploited in the more
recent work utilizing 2-bromoisobutyryl bromide for the
design of triblock copolymers with PFS blocks as outlined in
Scheme 1. A number of PEGs with Mns from 2 to 10 Â 103
(7, 11, 14) were employed as listed in Table 2. Additionally,
a liquid, dihydroxy terminated random copolymer (17) of
ethylene and propylene oxide (PEGPG with 74 wt.% PEG)
with a molecular weight of nearly 10,000 was also used.
In all cases the quantitative conversion of the hydroxyl
groups was verified by 1H NMR spectroscopy. Furthermore,
size exclusion chromatography (SEC) strongly implies that
the rather low PDI (Mw/Mn) from the polyethers is preserved
in the MIs (8, 12, 15, 18). The ATRP with FS allowed PFS
blocks, on the order of 39–84 wt.%, to be added to the
polyether blocks (9, 10, 13, 16, 19) as determined by a
2.3. Fluorinated star-shaped polymers
A tetrafunctional MI, 21, was prepared by reacting the
star-shaped PEG, 20 (based on a pentaerythritol core) shown
in Scheme 2 with 2-bromoisobutyryl bromide. Star-shaped
block copolymers, star-PEG-b-PFS (22), with flanking outer
PFS blocks were prepared by application of 21 for ATRP of
FS (Scheme 2, Table 3).
The SEC traces of both star-shaped block copolymers
(22) are completely shifted from those of 20 and 21
indicating that block copolymers were prepared. Based on
the SEC calibration with PEG standards and Mn of 21, the
Scheme 2. The principle route for preparation of fluorinated amphiphilic star-shaped block copolymers by ATRP.

K. Jankova, S. Hvilsted / Journal of Fluorine Chemistry 126 (2005) 241–250
245
Table 3
Exploitation of multifunctional ATRP initiators initially
based on the hydroxyl functions has been extended further to
include compounds that would provide principally spherical
growth to form star-shaped block copolymers. Scheme 3
illustrates the strategy starting from dipentaerythritol (23)
that is quantitatively converted to a hexafunctional ATRP
initiator (24) by the synthetic approach developed for PEG.
The remarkable symmetry of this rather large molecule, 24,
Composition and molecular weights of amphiphilic tetra-arm star polymers
Polymer
Composition
Molecular weights
b
b
a
PEG
PFS
Mn
Mw/Mn
Mn
(wt.%)a (wt.%)a
Star-PEG (20)
Star-PEG-b-PFS (22-1) 10
Star-PEG-b-PFS (22-2)
83
940 1.04
5700 1.30
6800 1.30
88
91
7300
9000
8
a
By 1H NMR.
1
is reflected by the simplicity of both the H and 13C NMR
b
By SEC with PEG calibration.
spectra [39]. Furthermore, a SEC analysis performed on an
OligoSEC column dedicated to relatively small macro-
molecules strongly implies that 24 is a pure compound.
24 was employed as initiator in a bulk ATRP of FS at
110 8C as shown in Scheme 3. Achieving high conversion of
FS by ATRP with 24 in bulk is not a problem, however, the
SEC traces are four to six modal and Mw/Mn is high. This is
probably caused by termination of each growing arm at
certain, but different stages due to the viscosity of the system
block copolymer compositions listed in Table 3 were
calculated by 1H NMR [38]. Even as little as 8–10 wt.% of
the hydrophilic part render the block copolymers (22)
amphiphilicity resulting in formation of micelles in THF/
water solutions. The micellization behaviour of (22) is under
investigation.
Scheme 3. The principal route for preparation of hexa-arm star fluoro and segmented block copolymers by ATRP.

246
K. Jankova, S. Hvilsted / Journal of Fluorine Chemistry 126 (2005) 241–250
Table 4
Molecular weights, glass transition temperatures, Tgs, and thermal stabilities of hexa-arm star-fluoropolymers of PFS prepared by ATRP
Polymera
Mn (RI)
Mw/Mn
Mn (LS)
Tg (8C)
TGAb first step (%)
TGAc residue (%)
25-1
25-2
25-3
25-4
25-5
25-6
25-7
25-8
25-9
25-10d
a
2600
4050
1.08
1.18
1.10
1.13
1.21
1.11
1.19
1.10
1.25
1.50
–
56
75
76
82
82
–
25.8
20.3
19.7
16.3
12.1
–
6.2
4.6
4.5
3.8
2.9
–
–
4400
–
5050
–
7200
–
8200
24000
–
10800
11400
21300
67400
87
–
11.9
–
2.5
–
27000
49500
99
99
–
–
2.2
1.1
Prepared by ATRP in bulk at 110 8C with the catalytic system CuBr/bipy using 0.03 mmol 24 and FS: 24 = 460 (25-1) for 2 min and (25-2 to 25-4) for
5 min; 1100 (25-5 to 25-7) and 1600 (25-8) for 7 min to 1–3% conversion.
b
Weight loss of the sample during the first step in the temperature range 277–350 8C.
Weight residue of the sample at 500 8C.
c
d
Prepared as above (see footnote a) in 10% xylene solution of FS to 18% conversion with FS: 24 = 1700 for 45 min.
at higher conversions. Different shapes of the SEC traces are
observed when the ATRP is performed either in bulk to only
10–20% conversion or in solutions of xylene, diphenyl ether
or THF. The polymers produced in this way show three
modal distributions with both a low and a high molecular
weight shoulder in addition to the main peak. The low
molecular weight shoulder is very small and presumably due
to an early termination of one or more growing arms of the
stars. While the high molecular weight analogue is probably
caused by star-star coupling, since the corresponding,
calculated Mn value is at least twice as high as the one
resulting from the main peak. Such a polymer with Mn of
67,400 and PDI = 1.50 is 25-10 listed in Table 4. It is
produced in 10% solution of FS in xylene to 18 wt.%
conversion and contains a slightly visible low molecular
weight shoulder and a high molecular weight peak
(Mn = 108,900; PDI = 1.28) with lower intensity than the
main peak with Mn of 43,100 and PDI = 1.04. Other attempts
to avoid these side reactions failed. Neither lower
temperatures (down to 85 8C) nor changing the ligand have
helped to increase the control. Nevertheless, a number of
well-defined hexa-arm star-PFSs with different, but increas-
ing Mns were prepared and presented in Table 4. They were
obtained by ATRP of FS in bulk at 110 8C to low conversions
(1–3%) employing the catalytic system CuIBr:bipy. Beyond
this limit of conversion for FS, irreversible recombination
reactions between stars became visible by SEC, while this
happened much later (at around 23% conversion) when
polymerizing St [39]. A SEC trace of the hexa-arm star-PFS
polymer (25-8) in THF prepared in this manner is provided
in Fig. 2. In this case the narrow peak shape of the SEC is
observed. Furthermore, we interpret the shape of the trace
and the low polydispersity index as an indication that all six
arms grew simultaneously at approximately the same rate
with little or no crosslinking due to radical coupling
reactions. However, star polymers of high functionality are
known to exhibit smaller radii of gyration and therefore
lower viscosity than linear ones. Thus, providing numbers
for Mn based on linear PS calibration appears meaningless.
The Mn figures reported in Tables 1 and 4 obtained by LS are
preliminary results that need further elaboration; however,
they seem to be more realistic values. The hexa-arm
fluoropolymers synthesized can be considered as fluorinated
nanoparticles. Thus, a novel route for synthesis of
fluorinated nanoparticles has been devised.
The Tg of the star-PFSs increases from 56 8C for sample
25-1 with Mn of 2600 to 87 8C for 25-7 with Mn of 10,800
and levels off at Mn of 21,300 (25-9) to 99 8C. A similar
dependence of Tg on molecular weight has previously been
observed by us for linear PFSs [30].
The arms of the star-PFSs are connected to the core via
ester linkages. In principle, this allows for cleavage by
hydrolysis which could provide a possibility to analyze the
size and molecular weight distribution of the cleaved linear
chains. Several of the star-PFSs prepared were subjected to
hydrolysis under basic conditions as in the case of star-PSs
[39]. Hydrolysis of the esters was attempted by refluxing a
THF solution of the star polymers with KOH (in ethanol)
[40,41]. To our surprise no cleavage was observed even after
several weeks of hydrolysis. This was the first indication that
Fig. 2. SEC of hexa-arm star-PFS (25-8).

K. Jankova, S. Hvilsted / Journal of Fluorine Chemistry 126 (2005) 241–250
247
Fig. 4. SEC of hexa-arm star-PS, 26-1 (—) and hexa-arm star-PS-b-PFS,
27-1 (– – –).
Fig. 3. TGA traces of the total weight loss of 23 (- - -), 24 (– Á – Á), 25-5 (—),
and 26 (– – –).
core scission as in case of star-PSs. Nevertheless, FS still
renders higher thermostability to the produced star polymers
as in the case of the linear ones [30].
the ester bonds are shielded and protected by the
hydrophobic FS and therefore inaccessible to hydrolysis.
Additionally, the H-containing groups in the core of the star-
The macroinitiator concept as the basis for a copolymer
was also pursued with the star polymers. However, the
fluoropolymer blocks should stay at the surface of the
nanoparticles in order to benefit mostly from the fluor-
opolymer characteristics. Therefore a hexa-arm star-PS (26)
was first prepared by use of the MI (24) in accordance with
the procedure outlined in Scheme 3. The SEC trace of the
isolated hexa-arm star-PS (26-1) is shown in Fig. 4. The
shape of the trace has strong resemblance to that of the hexa-
arm star-PFS presented in Fig. 2. The SEC analysis based on
PS calibration suggests a Mn of 11,600 (RI) and 16,700 (LS)
with Mw/Mn of 1.12–1.13.
1
PFSs could not be seen by H NMR.
In order to study further this fact the thermal stability of
the star-PFSs synthesized was investigated by thermogravi-
metric analysis (TGA) in a N2 atmosphere. Fig. 3 shows the
representative TGA curves of the total weight loss for 23, 24,
25-5, and a star-PS (26) with Mn = 10,000 (PDI = 1.13)
prepared by ATRP with the same MI (24). The thermo-
stability of 23 and 24 appears quite different. The
degradation of 23 starts at 260 8C, while the main
degradation of 24 already starts at 180 8C, and occurs in
two steps. The first step lasts until around 330 8C where the
resulting weight loss of 54–55%, seems to correspond to the
scission of the ester linkages in 24. The second step is
probably connected with degradation of condensation
species formed in the first step but with higher thermo-
stability than those of pure 23, as evidenced from the overlay
in Fig. 3. Also, all star-PFSs (25) investigated show two
steps of the weight loss—one starting at around 277–350 8C,
suggesting that a thermal degradation of the star polymers
involving the core happens prior to the main polymer itself.
This behaviour was additionally found for star-PSs, but the
first degradation step initiates almost 1008 lower, at around
180 8C. The extent of the first step for both 25 and 26
increases with decreasing Mn of the star polymer, which
correlates with the amount of the aliphatic part introduced
from 24. The weight losses of the star-PFSs samples in the
range 277–350 8C are presented in Table 4. At the same time
24 reveals a 9.4% weight residue at 500 8C while linear PFSs
or PSs produced by ATRP do not leave any residues [30]. For
example the weight loss from 277 to 350 8C decreases from
25.8 to 2.2% with increasing the Mn of the star-PFS from
2600 (25-1) to 67,400 (25-10), respectively, while the
amount of the residue at 500 8C decreases accordingly from
6.2 to 1.1 (Fig. 3, Table 4). These observations reveal that the
thermal degradation of star-PFSs involves as a first step the
It should be noted that although the SEC analysis is
performed with PS calibration we only consider the Mn value
indicative due to the spherical shape. On the other hand, the
polydispersity index of 1.12 is again believed to reflect the
controlled growth of the six PS arms with virtually no chain
coupling. The isolated hexa-arm star-PS (26-1) was then
used as a hexafunctional MI for a new ATRP procedure with
FS. The overlayed SEC trace of one of this block copolymer,
27-1, in Fig. 4 strongly suggests within the reliability of the
method that also the second PFS layer is grown in a
controlled fashion. In this way the PS core blocks of the
Table 5
Compositions and molecular weights of hexa-arm star-PS and segmented
star-PS-b-PFS copolymers prepared by ATRP
Polymer
Mn (PDI)
Mn (PDI)
PFS (wt.%)
(NMR)
Mn,NMR
(RI)
(LS)
26-1
27-1
27-2
26-2
27-3
26-3
27-4
11600 (1.12)
14300 (1.27)
20200 (1.27)
12400 (1.12)
15500 (1.14)
19200 (1.17)
33200 (1.14)
16700 (1.13)
0
71
89
0
–
40000
105000
–
33200 (1.15)
–
–
–
64
0
34500
–
34000 (1.18)
–
–
–

248
K. Jankova, S. Hvilsted / Journal of Fluorine Chemistry 126 (2005) 241–250
hexa-arm star-PS (26) could be surrounded by PFS blocks in
a corona, thus forming segmented, block copolymers. As PS
and PFS show similar glass transition temperatures
(ꢀ100 8C [30]) for the segmented block copolymer, 27-1,
only one Tg of 91 8C with a broad transition was found,
where the MI 26-1, itself had Tg of 82 8C. Table 5 lists 2
other star-PS MIs, 26-2 and 26-3, and the corresponding star
block copolymers 27-3 and 27-4 prepared in this manner.
core scission, PFS still renders higher thermal stability to the
produced star polymers as in the case of the linear analogues.
4. Experimental
4.1. Materials
Styrene, (Sigma–Aldrich) and 2,3,4,5,6-pentafluorostyr-
ene, FS (Fluorochem Limited, UK) were liberated from the
stabilizer and vacuum distilled before use. CuBr, 2,20-
bipyridine (bipy) and 1,4-dibromoxylene (all from Sigma–
Aldrich) were employed as received. THF and triethylamine
(TEA) were distilled from CaH2. The polyethers and glycols
were dried either by azeotropic distillation with toluene
(1H,1H0,10H,10H0-perfluorodecane-1,10-diol from ABCR-
Germany; star-PEG, PP150 from Perstorp Polyols, Sweden)
or at 180 8C for several hours (dipentaerythritol, from
Perstorp Polyols, Sweden).
3. Conclusions
We have attempted to demonstrate the versatility of
ATRP to enable preparation of different, novel fluorinated
polymer architectures based on the fully phenylfluorinated
styrene monomer, FS. The attractive ability of the resulting
PFS homopolymers to function as macroinitiators was
further exemplified by the facile preparation of linear
diblock copolymers with PS. Also triblock copolymers were
prepared by use of 1,4-dibromoxylene as the difunctional
initiator. Thus, varying amounts of FS can easily be
incorporated in the copolymers when used in combination
with St. A short, fluorinated dibromoester could be used as
ATRP initiator for St. The fluorinated copolymers enrich the
surface of thin films resulting in an increased advancing
water contact angle of 1078 (for PFS-b-PS) and 1008 (for PS-
b-F16-b-PS). These, as well as our former results with 4-
substituted fluoroalkoxy side chains PFSs (CA of 117–1228
[34]) strongly suggest that the fluorinated blocks in the
investigated fluorinated copolymers of PS migrate to the
open film surface and create low surface energy films.
Multiblock copolymers based on different difunctional
bromine initiators, e.g. 1,4-dibromoxylene, were likewise
synthesized. Even more important is the possible quanti-
tative bromoester derivatization of compounds with two or
more hydroxyl groups. Thus, the short fluorinated dibro-
moester gave rise to a pentablock copolymer with
alternating fluorinated and PS blocks. Various hydroxyl-
terminated polyethers of different lengths were converted to
macroinitiators according to this concept and a number of
triblock copolymers with PFS end blocks were prepared.
These amphiphilic block copolymers have very interesting
material properties such as good Li+ complexation while
preserving excellent film forming capability. Since the Li+
conductivity is high, such materials seem interesting for
solid-state electrolyte applications in batteries. The flex-
ibility of the initiator concept could be broadened to the
synthesis of a tetrafunctional hydrophilic MI based on a star-
PEG, which was extended to amphiphilic star-PEG-b-PFS
copolymers. A hexafunctional MI with a dipentaerythritol
core was the basis for preparation of a number of
hydrophobic hexa-arm star fluoropolymers. Thereby a route
for preparation of novel fluorinated nanoparticles has been
devised. These nanoparticles may consist entirely of the
fluoropolymer PFS or have a PS core surrounded by a PFS
corona. Even though the thermal degradation involves first a
4.2. Synthesis
The dried polyols were converted to MI by reaction with
2-bromoisobutyryl bromide catalyzed by TEA in anhydrous
THF [36]. In a typical ATRP procedure [30] an appropriate
amount of the MI was dissolved either in the monomer or in
xylene in a Schlenk tube and the necessary amounts of FS,
CuBr, and bipy in order to reach Mn,target (Mn,target = [FS]0/
[MI]0 at 100% conversion) were added. The molar ratio of
initiating groups:CuBr:bipy was kept 1:1:2. After three
freeze-thaw cycles, the polymerization tube was immersed
in a preheated oil bath at 110 8C. The polymerization was
stopped after the appropriate time by transferring the tube to
a cooling bath. The polymer solutions were then filtered, if
necessary after dilution with THF, and precipitated in
methanol. When the ATRP’s of FS were run until low
conversions the big excess of the monomer was vacuum
distilled prior to dilution with THF and precipitation.
Conversions of FS in the homo- and block copolymers
formed were determined gravimetrically after vacuum
drying.
4.3. Analysis: NMR spectroscopy
The macroinitiators and the block copolymers were
characterized by 1H NMR, using a Bruker 250 MHz
spectrometer and DMSO-d6 or CDCl3 as solvents. 19F
NMR in C6F6 constitutes only of three broad peaks for the
di- and triblock copolymers of PFS and PS at À154.6 (2Far),
À147.6 (1Far) and À137.0 (2Far) ppm as for the PFS
homopolymer. The composition of the various block
copolymers of PFS with PS or polyethers (Tables 1 and
5) was determined therefore by 1H NMR in a similar way as
published before [30,38]. Ratio analysis of the area of
aliphatic protons (from both PFS and PS main chains, 1.8–
3.3 ppm) and the area of the aromatic protons (from only PS,

K. Jankova, S. Hvilsted / Journal of Fluorine Chemistry 126 (2005) 241–250
249
6.5–7.4 ppm) was used. The content of PFS, wPFS (in wt.%),
in the block copolymer with PEG presented in Table 2 was
Acknowledgements
calculated by wPFS = 194APFS/(194APFS + 33APEG), where
PFS is the area of the aliphatic protons of the PFS segment
Aage and Johanne Louis-Hansen’s Foundation (DK) and
Materials Research, Danish Research Agency is gratefully
acknowledged for financial support. We additionally thank
Sokol Ndoni, Risø National Laboratory, DK, for performing
the SEC light scattering (LS) analyses. Perstorp Polyols, SE,
is acknowledged for kindly supplying dipentaerythritol and
PP150.
A
between 1.8 and 2.9 ppm and APEG is the area of the PEG
protons around 3.6 ppm. The amount of PFS in the PEGPG
block copolymers was determined in a similar way. The
composition of the star-PEG, PP 150, is calculated from the
mass of the 15 ethylene oxide units as compared to the total
mass of 800 g molÀ1, given by the producer. The content of
ethylene oxide in mol% in the amphiphilic PP 150 based
star-shaped block copolymers is calculated by the ratio of
the area for the 15 PEG units (60 protons) at 3.2–4.1 ppm as
compared with its sum of the area for the PFS protons (1.5–
2.9 ppm).
References
[1] J. Nieratschker, Kunststoffe 89 (1999) 130–134.
[2] H.-J. Le, M.-H. Lee, M.C. Oh, J.-H. Ahn, S.G. Han, J. Polym. Sci., Part
A: Polym. Chem. 37 (1999) 2355–2361.
The Mn,NMR of the polyether based block copolymers was
calculated as:
[3] C. Pitois, S. Vukmirovic, A. Hult, D. Wiesmann, M. Robertsson,
Macromolecules 32 (1999) 2903–2909.
[4] J.-P. Kim, W.-Y. Lee, J.-W. Kang, S.-K. Kwon, J.-J. Kim, J.-S. Lee,
Macromolecules 34 (2001) 7817–7821.
polyether
n;SEC
M
Mn;NMR
¼
þ150fn;
[5] C. Pitois, D. Wiesmann, M. Lingren, A. Hult, Adv. Mater. 13 (2001)
1483–1487.
ð100 À wPFSÞ=100
[6] L.M. Han, R.B. Timmons, W.W. Lee, Y.C. Chen, Z. Hu, J. Appl. Phys.
84 (1998) 439–444.
´
[7] P. Gavelin, P. Jannasch, B. Wesslen, J. Polym. Sci., Part A: Polym.
while the Mn,NMR of the amphiphilic star-shaped block
copolymers was determined as:
Chem. 39 (2001) 2223–2232.
´
[8] P. Gavelin, P. Jannasch, I. Furo, E. Pettersson, P. Stilbs, D. Topgaard,
660
Mn;NMR
¼
þ150fn þ 132;
wPEG=100
¨
O. Soderman, Macromolecules 35 (2002) 5097–5104.
[9] T.M. Chapman, R. Benrashid, K.G. Marra, J.P. Keener, Macromole-
cules 28 (1995) 331–335.
where 150 is the mass of the initiating Br-ester end group,
660 is the mass of 15 ethylene oxide units, wPEG is the wt.%
of 15 ethylene oxide units, and fn is the number-average
functionality of the starting polyether (2 for PEGs and
PEGPG, and 4 for star-PEG).
[10] T.M. Chapman, K.G. Marra, Macromolecules 28 (1995) 2081–2085.
[11] D.R. Iyengar, S.M. Perutz, C.-A. Dai, C.K. Ober, E.J. Kramer,
Macromolecules 29 (1996) 1229–1234.
[12] J. Wang, G. Mao, C.K. Ober, E.J. Kramer, Macromolecules 30 (1997)
1906–1914.
[13] J. Wang, C.K. Ober, Macromolecules 30 (1997) 7560–7567.
¨
[14] A. Boker, K. Reihs, J. Wang, R. Stadler, C.K. Ober, Macromolecules
4.3.1. SEC analysis
33 (2000) 1310–1320.
Molecular weights were determined by SEC employing a
Viscotek 200 instrument equipped with a PLguard and 2
PLgel mixed D columns in series from Polymer Laboratories
(PL). Measurements were performed in THF at room
temperature with a 1 mL minÀ1 flow and RI detection;
molecular weights were calculated using either PS in the
range 7 Â 102 to 4 Â 105 or PEG narrow molecular weight
standards (102 to 6 Â 104) from PL and the TriSECTM
Software.
[15] K. Sugiyama, T. Nemoto, G. Koide, A. Hirao, Macromol. Chem. 181
(2002) 135–153.
[16] A. Hirao, G. Koide, K. Sugiyama, Macromolecules 35 (2002) 7642–
7651.
[17] K. Li, P. Wu, Z. Han, Polymer 43 (2002) 4079–4086.
[18] L. Andruzzi, E. Chiellini, G. Galli, X. Li, S.H. Kang, C.K. Ober, J.
Mater. Chem. 12 (2002) 1684–1692.
¨
[19] A. Boker, T. Herweg, K. Reihs, Macromolecules 35 (2002) 2937–
4929.
[20] T. Miyata, H. Yamada, T. Uragami, Macromolecules 34 (2001) 8026–
8033.
¨
[21] H. Hussain, H. Budde, S. Horing, K. Busse, J. Kressler, Macromol.
4.3.2. DSC analysis
Chem. Phys. 203 (2002) 2103–2112.
Thermal analyses were executed with a differential
scanning calorimeter DSC Q1000 from TA Instruments in a
temperature range of À100 to 200 8C at a heating rate of
10 8C minÀ1 under nitrogen. The glass transition tempera-
ture (Tg) was determined automatically by the instrument
from the second heating trace and is reported as the midpoint
of the thermal transition.
[22] H. Hussain, K. Busse, J. Kressler, Macrolmol. Chem. Phys. 204 (2003)
936–946.
[23] K.T. Lim, M.Y. Lee, M.J. Moon, G.D. Lee, S.-S. Hong, J.L. Dickson,
K.P. Johnston, Polymer 43 (2002) 7043–7049.
´
[24] S. Perrier, S.G. Jackson, D.M. Haddleton, B. Ameduri, B. Boutevin,
Macromolecules 36 (2003) 9042–9049.
´
[25] B. Ameduri, B. Boutevin, Well-Architectured Fluoropolymers:
Synthesis, Properties and Applications (F-Telomers, Telechelics,
and Alternated, Block and Graft Copolymers), Elsevier, Amsterdam,
2004.
Thermal degradation was investigated by TGA per-
formed with a TGA Q500 from TA Instruments recording
the total weight loss on approx. 10–12 mg samples from
room temperature to 600 8C at a rate 10 8C minÀ1 in a
[26] P. Lacroix-Desmazes, B. Boutevin, D.K. Taylor, J.M. DeSimone,
Polym. Preprints 43 (2) (2002) 285–286.
[27] V. Coessens, T. Pintaur, K. Matyjaszewski, Prog. Polym. Sci. 26
(2001) 337–377.
nitrogen flow of 90 mL minÀ1
.

250
K. Jankova, S. Hvilsted / Journal of Fluorine Chemistry 126 (2005) 241–250
[28] K. Matyjaszewski, J. Xia, Chem. Rev. 101 (2001) 2921–2990.
[29] J. Qiu, K. Matyjaszewski, Macromolecules 30 (1997) 5643–5648.
[30] K. Jankova, S. Hvilsted, Macromolecules 36 (2003) 1753–1758.
[31] J.A. Johnson, C.S. Gudipati, K.L. Wooley, Polym. Preprints 44 (2)
(2003) 238.
[35] B. Gao, X. Chen, B. Ivan, J. Kops, W. Batsberg, Macromol. Rapid
Commun. 18 (1999) 1095–1100.
[36] K. Jankova, X. Chen, J. Kops, W. Batsberg, Macromolecules 31 (1998)
538–541.
[37] K. Jankova, J.H. Truelsen, X. Chen, J. Kops, W. Batsberg, Polym. Bull.
42 (1999) 153–158.
[32] S. Hvilsted, S. Borkar, L. Abildgaard, V. Georgieva, H.W. Siesler, K.
Jankova, Polym. Preprints 43 (2) (2002) 26–27.
[38] K. Jankova, P. Jannasch, S. Hvilsted, J. Mater. Chem. 14 (2004) 2902–
2908.
[33] S. Hvilsted, S. Borkar, H.W. Siesler, K. Jankova, Novel fluorinated
polymer materials based on 2,3,4,5-tetrafluoro-4-methoxystyrene, in:
K. Matyjaszewski (Ed.), Advances in Controlled/Living Radical
Polymerization, ACS Symposium Series, vol. 854, American Chemi-
cal Society, Washington, DC, 2003, pp. 236–249 (Chapter 17).
[34] S. Borkar, K. Jankova, H.W. Siesler, S. Hvilsted, Macromolecules 37
(2004) 788–794.
[39] K. Jankova, M. Bednarek, S. Hvilsted, J. Polym. Sci., Part A: Polym.
Chem., submitted for publication.
[40] S. Angot, K.S. Murthy, D. Taton, Y. Gnanou, Macromolecules 31
(1998) 7218–7225.
[41] A.P. Narrainen, S. Pascual, D.M. Haddleton, J. Polym. Sci., Part A:
Polym. Chem. 40 (2002) 439–450.