Low-fluorinated homopolymer from heterogeneous ATRP of 2,2,2-trifluoroethyl methacrylate mediated by copper complex with nitrogen-based ligand

Article abstract of DOI:10.1016/j.jfluchem.2011.05.027

We report the preparation of low-fluorinated homopolymer via heterogeneous atom transfer radical polymerization (ATRP) of 2,2,2-trifluoroethyl methacrylate (TFEMA) using 2,2′-bipyridine (bpy), N,N,N′,N′,N″- pentamethyldiethylenetriamine (PMDETA), and tris(2-(dimethylamino)ethyl)amine (Me₆TREN) as representatives for di-, tri-, and tetradentate amine ligands, respectively. The ATRP was better controlled, yielding polymers with controlled molecular weights and low polydispersities (M_w/M _n ca. 1.11) when bpy was used as a ligand than when PMDETA was used. This was further supported by the results of our kinetic and chain extension studies. However, the ATRP of TFEMA had lower monomer conversions and gel formation when Me₆TREN was used as the ligand. Further reported are the thermal-properties, as well as the surface properties of the films from the resulting polymers with different molecular weights.

Full text of DOI:10.1016/j.jfluchem.2011.05.027

Journal of Fluorine Chemistry 132 (2011) 562–572
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Low-fluorinated homopolymer from heterogeneous ATRP of 2,2,2-trifluoroethyl
methacrylate mediated by copper complex with nitrogen-based ligand
a
a
a
a
Guping He a, Ganwei Zhang a, Jiwen Hu a, , Jianping Sun , Shenyu Hu , Yinhui Li , Feng Liu ,
*
Dingshu Xiao a, Hailiang Zou a, Guojun Liu b
a Guangzhou Institute of Chemistry, Chinese Academy of Sciences, Guangzhou 510650, PR China
b Department of Chemistry, Queen’s University, 90 Bader Lane, Kingston, Ontario, Canada K7L 3N6
A R T I C L E I N F O
A B S T R A C T
Article history:
We report the preparation of low-fluorinated homopolymer via heterogeneous atom transfer radical
polymerization (ATRP) of 2,2,2-trifluoroethyl methacrylate (TFEMA) using 2,20-bipyridine (bpy),
N,N,N0,N0,N00-pentamethyldiethylenetriamine (PMDETA), and tris(2-(dimethylamino)ethyl)amine
(Me6TREN) as representatives for di-, tri-, and tetradentate amine ligands, respectively. The ATRP
was better controlled, yielding polymers with controlled molecular weights and low polydispersities
(Mw/Mn ca. 1.11) when bpy was used as a ligand than when PMDETA was used. This was further
supported by the results of our kinetic and chain extension studies. However, the ATRP of TFEMA had
lower monomer conversions and gel formation when Me6TREN was used as the ligand. Further reported
are the thermal-properties, as well as the surface properties of the films from the resulting polymers
with different molecular weights.
Received 8 April 2011
Received in revised form 20 May 2011
Accepted 27 May 2011
Available online 29 June 2011
Keywords:
Nitrogen-based ligand
ATRP
2,2,2-Trifluoroethyl methacrylate
Heterogeneous
ß 2011 Elsevier B.V. All rights reserved.
1. Introduction
zation, conventional free radical polymerization, etc. [1]. Among
them, conventional free radical polymerization of fluoromonomer,
Fluorinated (co)polymers, which contain fluorine in their
backbones or side chains, have attracted continuous attention
from researchers engaged in both fundamental and industrial
research [1], since the invention of polytetrafluoroethylene (PTFE)
in 1938 [2] and the development of soluble perfluoropolymer
(Teflon@AF) in 1992 [3]. Indeed, owing to the strong electronega-
such as fluorinated (meth) acrylates, styrenic and alkenes [15–17],
using multiple polymerization processes, e.g. bulk, solution,
emulsion, or precipitation polymerization methods [18,19]. An
evident drawback of conventional free radical polymerization,
however, is that it generates (co)polymers with uncontrolled
molecular weights and molecular weight polydispersity (Mw/Mn),
as well as ill-defined architecture, because of the very high
nonselective activity of the radical, the high termination and
transfer of the propagating radical species [18]. To overcome these
disadvantages, new techniques are developed based on either
reversible deactivation of polymer radicals or a degenerative
transfer process, called ‘living’ or controlled radical polymerization
(CRP). The outstanding achievements of CRP methods in the past
decade allow for the development of advanced well-defined
(co)polymers with various architectures (i.e. telechelic, block,
graft, or star copolymers) having predictable molecular weights
and low molecular weight polydispersities. As a consequence, the
area of self-assembly has also benefited from this, with advances in
the preparation of unique structures and many functional
materials from self-assembly of copolymers with well-defined
structures [20,21].
˚
tivity and small van der Waals radius (1.32 A) of fluorine atom, as
well as the strong C–F bond (a high dissociation energy of 485 kJ/
mol), these kinds of polymers possess unique combined properties.
They have good biocompatibility, high thermal stability, good
chemical resistance, superior weatherability, oil and water
repellence, low flammability, as well as low refractive index,
etc. [1,4–7]. Because of these, they have a diverse range of
applications in the preparation of many functional materials with
notable properties, such as biomaterials, [8] surfactants, [9]
lubricants, [10] insulators, [11] ion conducting materials for
lithium-ion batteries, [12] proton conducting membranes for fuel
cells, paints [13] and coatings [14].
Normally, the incorporation of fluorine into (co)polymers can
be accomplished by
a variety of synthetic polymerization
techniques including cationic polymerization, anionic polymeri-
So far, there have been a number of reports dealing with the
preparation of fluorinated or per-fluorinated (co)polymers via
controlled/living radical polymerization. Examples include revers-
ible addition-fragmentation chain transfer (RAFT) polymerization,
* Corresponding author. Fax: +86 20 85232307.
E-mail address: hjw@gic.ac.cn (J. Hu).
0022-1139/$ – see front matter ß 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2011.05.027

G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
563
atom transfer radical polymerization (ATRP), nitroxide-mediated
polymerization (NMP), iodine transfer polymerization (ITP), as
well as CRP controlled by boron derivates as activated by oxygen.
These controlled radical polymerization techniques have been
discussed in detail in a recent book and comprehensive reviews
[22]. Among these CRP techniques, ATRP has become one of the
most efficient and widely used methods to obtain (co)polymers
with different topologies [23,24] since the independently pio-
neered work by Sawanmato [25a] and Matyjaszewski in 1995
[25b]. In general, this technique comprises a halide functionalized
initiator, a transition metal ion (normally, copper(I) ion), and a
ligand which forms a complex with the metal ion [25].
dioxide [41], and butyl methacrylate and TFEMA were copoly-
merized via RAFT miniemulsion polymerization [42]. To date,
there have also been a few reports on the ATRP TFEMA [3–47].
For example, Perrier reported the synthesis of PTFEMA with
average polymerization degrees less than 100 at 90 8C. 2-ethyl
bromoisobutyrate and Cu(I)Br/N-(n-pentyl)-2-pyridylmethani-
mine in toluene were used as the initiating system. The resultant
polymers had relatively low polydispersity at all stages of
polymerization (PDI ꢀ 1.30). They further demonstrated the
poor initiator efficiency and a gradual loss of reactive species
resulting from irreversible termination during polymerization
[43]. Hvilsted and his coworkers reported the controlled ATRP of
TFEMA initiated with 2-ethylbromoisobutyrate (2EBiB) and
Ligands in the ATRP are employed to ensure sufficient
solubilization of the transition-metal salt in organic medium
and to tune the proper reactivity and dynamic halogen exchange
between the metal center and the dormant species or persistent
radical [25]. Most of the prior research on ATRP has focused on
the nitrogen-based ligands for copper-based ATRP [26], since
ligands based on sulfur, oxygen, or phosphorus are often more
expensive and less effective due to inappropriate electronic
effects or unfavorable binding constants compared to that of
nitrogen-based ligands [24a]. So far, two common types of
nitrogen ligand developed as active and efficient complexion
agents for copper-mediated ATRP [27] are (i) aromatic com-
pounds containing sp-hybridized nitrogen atoms available in the
form of R155N–R2 (such as 2,20-bipyridine (bpy) and its
derivatives) and (ii) aliphatic substances containing sp [2]
nitrogen atoms in the form of R1–NR2–R3 (such as N,N,N0,N0,N00-
pentamethyldiethylenetriamine (PMDETA) and tris(2-(dimethy-
lamino)ethyl)amine (Me6TREN) as representatives for tri- and
tetradentate amine ligands frequently used for ATRP). Aromatic
Cu(I)Br/N-(n-propyl)-2-pyridylmethanimine
to
synthesize
amphiphilic block copolymers of TFEMA and MMA, 2-methox-
yethyl acrylate, or poly(ethylene oxide) methyl ether methac-
rylate [44]. Zhu and coworkers reported the preparation and
aggregating behavior of a ABCBA-type pentablock copolymer,
PDMAEMA-b-PTFEMA-b-PCL-b-PTFEMA-b-PDMAEMA via conse-
cutive ATRP using poly(e-caprolactone), Br–PCL–Br, as the
initiating block to polymerize TFEMA in the presence of CuCl/
bpy in DMF at 85 8C [45]. A series of well-defined diblock
copolymers of acrylic acid with partially fluorinated acrylate
(TFEMA) and methacrylate monomers were synthesized using
ATRP to serve as potential 19F MRI imaging agents by Whittaker
and colleague. They declared that poly(tert-butyl acrylate)-b-
poly[butyl acrylate-co-FEMA] (ptBA-b-p(BA-co-FEMA)) copoly-
mers were initiated from macroinitiator, PtBA–Br, at 90 8C in
toluene using CuBr/PMDETA as catalytic system, and the
conversion limited to 40% for the second block [46]. Wooley
and her coworkers also prepared amphiphilic hyperbranched
partially fluorinated copolymers for the potential application as
19F MRI imaging agents [47]. The work on ATRP of TFEMA
reported, so far, indicates either lower monomer conversion or
loss of control, therefore, it is worthy to carry a systematically
study on ATRP of TFEMA.
nitrogen-based ligands usually induce
a higher oxidation
potential of the complexed metal center, leading to a greater
tendency toward radical deactivation in the atom transfer
equilibrium and therefore slower overall polymerization rates
[28–30]. On the other hand, tridentate and tetradentate
aliphatic amine ligands generally provide faster polymerizations
without causing significant broadening of the molecular weight
distribution of the product [28]. In other words, the use of an
appropriate ligand is still of major importance to provide an
efficient halogen-exchange reaction between the dormant and
the active species in the ATRP as well as to produce polymers
with controlled molecular weights and low polydispersities [31].
2,2,2-Trifluoroethyl methacrylate (TFEMA), a commercially
available monomer with a low level of fluorination, generates
the (co)polymers with combined features as those from typical
methacrylate monomers and fluorine-containing monomer simul-
taneously [32]. Therefore, the (co)polymers derived from TFEMA
can be utilized for the fabrication of many functional materials. For
example, they can be used in coatings, resin modification, and
adhesives because of their non-cohesiveness as well as water and
oil repellency [33–35]. They are utilized in contact lenses because
of enhanced oxygen permeability in a fluorinated matrix [36].
Furthermore, their electron-withdrawing properties have facili-
tated their use as an electrical charge control agent in photocopy
toners [37], and their superior mechanical strength and low
refractive index have facilitated their use in optical fiber [38]. Also,
the fluorine electronic effect is taken advantage of in 19F MRI
imaging applications [39].
1.1. TFEMA ATRP
When one of the oxidation states of the metal–ligand
complex is less soluble, a heterogeneous system can originate,
which can result in complex polymerization kinetics and even
loss of polymerization control. This can be usually found in
systems with bpy [48,49] and multi-dentate aliphatic amine
ligands [50]. However, these disadvantages are accompanied by
some major advantages, namely the ligands are cheap, easily
accessible (especially the multi-dentate amines), and easily
separated from the polymer solution because of the heteroge-
neity of the metal/ligand complexes [51]. In the present
contribution, in an effort to have better control on TFEMA
ATRP, we study the heterogeneous ATRP of TFEMA using bpy,
PMDETA, and Me6TREN as representatives for di-, tri-, and
tetradentate amine ligands, respectively (Fig. 1). The properties
of the resulting homopolymers, PTFEMA, with various molecular
weights are also investigated.
The controlled polymerization of TFEMA could be dated to
2001, in which Roussel and Boutevin reported the controlled
polymerization of TFEMA by the ‘‘iniferter (initiation-transfer-
termination)’’ method using a substituted fluorinated tetra-
phenylethane-type initiator [40]. More recently, RAFT (co)po-
lymerizations involving TFEMA were reported. For example,
PTFEMA was grafted onto ramie fibers in supercritical carbon
Fig. 1. Ligands used for ATRP of TFEMA.

564
G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
2. Experimental
2.1. Materials and reagents
the product was obtained as colorless oil at a yield of 89%. 1H NMR
(CDCl3, 400 MHz): 2.20 (s, 18H, CH3); 2.3–2.37 (m, 6H, CH2); d
d
d
2.56–2.60 (m, 6H, CH2).
TFEMA, purchased from Harbin Xeogia Fluorine-Silicone
Chemical Company, China, was washed with 2 wt.% aqueous
solution of sodium hydroxide, and then with double-distilled
water several times till neutralization. The organic layer was
collected and dried over anhydrous magnesium sulfate before
distillation under reduced pressure just before use. Copper(I)
bromide (CuBr, Fluka, 99+%) was purified by stirring in glacial
acetic acid at 80 8C for over 8 h, followed by washing with dry
methanol over 10 times before drying in vacuum at room
temperature for 48 h, according to the method reported in our
previous paper [52]. Copper(I) chloride (CuCl, Aldrich, 99%) was
purified using a method similar to that used for CuBr. N,N,N0,N0,N00-
pentamethyldiethylenetriamine (PMDETA, Aldrich, 99+%) and 2,20-
bipyridine (bpy) (Aldrich, 99%) were used as received. Cyclohexa-
none purchased from Tianjin Baishi Chemical (99.8%) was initially
decolored by active carbon, and then stirred with calcium hydride
(Aldrich, 99%) overnight before distillation under reduced pressure
prior to use. Tetrahydrofuran (THF), diethyl ether, acetonitrile and
hexane were all of analytical grade and distilled over sodium wires
prior to use. Hydroxyethyl methylacrylate (HEMA) was purified
using the following procedures: washing the aqueous solution of
HEMA (100 mL HEMA and 300 mL deionized water) with hexanes
for eight times (8ꢁ 50 mL), after removing the organic layer, salting
HEMA out from the aqueous phase by addition of NaCl, drying over
anhydrous magnesium sulfate, and finally distilling under reduced
pressure. Trimethylsiloxyethyl methacrylate (HEMA-tms) was
prepared and purified according to a literature procedure [53].
Other reagents, if not specified, were treated using normal
procedures that were used in the lab.
2.4. Characterization
1H NMR spectroscopy measurements were recorded on a Bruker
DMX-400 spectrometer with Varian probe in deuterated
a
chloroform. The concentration of sample in CDCl3 typically used
for 1H NMR measurement was about 5–10 mg/mL. Size exclusive
chromatography (SEC) was performed using a Waters 1515 series
GPC system equipped with a Waters 2414 refractive index (RI)
detector and a set column of styragel HR4 and HR3 at 35 8C. The
correlation between the elution time and the polymer molecular
weight was predetermined by low-polydisperse polystyrene
standards. The samples usually at a concentration of 5–10 mg/
mL in DMF were filtrated through 0.45
mm before injection. The
mobile phase was HPLC grade DMF at a speed of 0.6 mL/min. FT-IR
spectra were recorded using a Bruker FT-VERTEX 70 at a scan range
of 400–4000 cmꢂ1 with an accuracy of 2 cmꢂ1. Sample for testing
was initially mixed with KBr and pressed to a circular sheet before
FTIR evaluation. The glass transition temperatures (Tg) were
determined with a DSC Q200 system (TA Co. Ltd., USA). Sample
was heated from room temperature to 150 8C at a rate of 10 8C/min
and kept for 3 min before cooled to ꢂ50 8C at a rate of 10 8C/min to
remove any effects induced by prior treatments. The Tg was then
determined by consecutive heating from ꢂ50 8C to 200 8C at a rate
10 8C/min. Thermogravimetric analysis (TGA) was performed on a
thermoanalyzer system (model Q600SDT, TA Co. Ltd., USA) by
heating 5–8 mg of sample from room temperature to 500 8C at a
rate of 20 8C/min under nitrogen atmosphere. The onset tempera-
ture, decomposition temperature and residual mass were calcu-
lated with the TA Instruments Universal Analysis 2000 software.
All samples were vacuum dried at 50 8C for 24 h prior to
measurements. Contact angle measurements were estimated on
an OCA40 plus contact angle system apparatus from Dataphysics
with contact angles attained by the drop-shape (geometry)
method [54]. Films for contact angle measurements were prepared
at room temperature under dry atmosphere in a desiccator by
dropping a few drops of polymer solution in THF at 0.2 g/mL onto
glass slide, which was initially cleaned in sequence with methanol,
2.2. Preparation of methoxyl ethylene 2-bromoisobutyrate (MEBrIB)
MEBrIB was synthesized by reacting anhydrous ethylene glycol
monomethyl ether with 1.2 equivalent of 2-bromoisobutyryl
bromide in dry diethyl ether in the presence of triethylamine as
acid absorbent at room temperature overnight. After removing
triethylamine hydrogen bromide salt by centrifugation, the
reaction mixture was successively washed with 2 N HCl, saturated
sodium carbonate, and double distilled water before drying over
anhydrous magnesium sulfate. The crude product was isolated as a
slight colorless liquid upon removal of the solvent and purified by
distillation under vacuum (ꢀ1 mmHg) at 74–76 8C at a yield of
acetone, and deionized water. A drop of testing liquid (5
ml, pure
water or CH2I2) was placed via syringe onto the resulting polymer
film and the image was immediately sent via the CCD camera to the
computer for analysis. Contact angle was reported as averaged at
least ten different positions for the same film.
90%. 1H NMR (CDCl3, 400 MHz):
3H); 1.93 (s, 6H).
d 4.3 (t, 2H), d 3.6 (t, 2H), 3.37 (s,
d
2.5. General procedures for ATRP of TEFMA
2.3. Preparation of tris(2-(dimethylamino)ethyl)amine (Me6TREN)
The ATRP of TFEMA was performed in cyclohexanone using
CuBr as a catalyst in the presence of different ligands. In a typical
run, cyclohexanone (6 ml), TFEMA (5.011 g, 29.76 mmol), MEBrIB
(50.3 mg, 0.2237 mmol), and bpy (69.8 mg, 0.4475 mmol) were
added in one of the flasks of a home-made two-flask set connecting
two 25 ml flasks via a glass pipe with a diameter of 0.8 cm and
length of 5 cm, while CuBr (32.1 mg, 0.2237 mmol) was loaded to
the other flask. The liquid mixture was bubbled with pure argon for
over half an hour before subjected to three freeze-pump-thaw
cycles, and then, carefully transferred to the flask containing CuBr
and put into an oil-bath with a constant temperature of 80 8C. The
reaction was stopped by freezing the flask with liquid nitrogen
before exposure to air as indicated by color of the solution turning
from pale green to blue within 1 min. Then, the blue reaction
mixture was diluted with 5 mL of THF before passing through a
neutral alumina column to remove copper complex. The filtrate
was concentrated to about 10 mL with rotary-evaporation before
Me6TREN was prepared as we previously reported [52]. Briefly,
to prepare the salt (ClNH3CH2CH2)3NHCl, 30 mL of 3.0 M HCl in
methanol was added dropwise to 4.0 mL or 0.027 mol of tris(2-
aminoethyl)amine in 50 mL of methanol. After stirring at room
temperature for 1 h, the precipitate was filtered and washed with
50 mL of methanol thrice to yield 6.72 g or 0.026 mol of product in
98% yield. Then, 6.72 g of (ClNH3CH2CH2)3NHCl, 10 mL of distilled
water, 50 mL of formic acid, and 46 mL of a formaldehyde aqueous
solution were mixed. The mixture was heated under stirring in a
120 8C oil bath for 6 h before volatile components were removed by
rotory-evaporation. To the solid residue was then added 100 mL of
10 wt.% NaOH aqueous solution. The resulting aqueous phase was
extracted with 100 mL of diethyl ether for four times. The organic
layer was collected, dried over anhydrous NaOH, and concentrated
by rotary evaporation. After vacuum distillation at 62 8C, 6.0 g of

G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
565
precipitated out over 100 mL of hexane. The crude product was
purified by repeatedly dissolving in 5 mL of THF and precipitated
out over 100 mL of hexane twice before dried overnight under
vacuum to generate 3.8 g of polymer as white powder. 1H NMR
(CDCl3):
–CH2);
d
4.32 (br s, –CH2CF3);
d 3.35 (br s, –OCH3); d 1.75–2.05 (m,
d
0.75–1.35 (m, –CH3); Mw/Mn = 1.11.
2.6. Kinetic study of ATRP of TFEMA
To monitor the ATRP kinetic, in situ 1H NMR spectra were
collected with different time intervals. In a representative run,
cyclohexanone (5 ml), TFEMA (5.0310 g, 29.8 mmol), MEBrIB
(66.9 mg, 0.298 mmol), and bpy (92.8 mg, 0.595 mmol) were
added to one flask of a two-connective flask set and CuBr (42.7 mg,
0.298 mmol) was added to the other flask of the connective flask.
The reactants were bubbled for half an hour with argon before
subjected to three freeze pump thaw cycles and then transferred
the liquid reactants to the flask containing CuBr. The solution was
put into an oil-bath preheated to 80 8C with constant stirring.
Samples were withdrawn with time intervals in order to
investigate the nature of the TFEMA polymerization by 1H NMR
and SEC analyses. Conversion and SEC analysis were directly
performed for the withdrawn samples. Conversions were estimat-
ed from 1H NMR analysis by comparison of sum of the peak
intensity for the double bond –CHdCH2 (5.85 and 6.19 ppm)
protons from TFEMA to the sum of that for –CH2–CF3 from
Fig. 2. FT-IR spectrum of homopolymer PTFEMA.
[L]0 represent the initial concentration of monomer, initiator, and
ligand, respectively. The reaction mixture turned from yellow to
green right away, and then the whole reaction mixture succes-
sively becomes dark green and opaque. The viscosity of the
reaction mixture progressively increased after 1 h indicating
polymerization [55]. After stopping the reaction by freezing the
reaction mixture with liquid nitrogen and introducing the air to the
reaction mixture, the green color of the mixture immediately
turned blue demonstrating that all the Cu(I) complex transferred to
Cu(II) complex [55]. The yield of the product as white powder was
about 69% after purification as judged by gravimetric method.
Fig. 2 shows the FT-IR spectrum of homopolymer PTFEMA from
ATRP. The characteristic band at 1751 cmꢂ1 corresponds to ester
carbonyl (iC55O) bonds, and the peaks at 660 cmꢂ1 and 1281 cmꢂ1
are attributed to the stretching and bending vibration of C–F bonds,
respectively [41]. The disappearance of absorption peak at
1640 cmꢂ1 from stretching vibration of C55C bond in the monomer
TFEMA indicates successful polymerization of TFEMA [56].
Fig. 3 shows 1H NMR spectrum and peak assignments of
initiator, monomer, and PTFEMA, respectively. The changes in the
chemical shift of CH2 (vinyl group in the monomer) protons from
6.20 and 5.67 ppm to 2.17 ppm (backbone in the polymer) indicate
the polymerization of monomer. The chemical shift belongs to
protons of methylene groups attached to –CF3 moved from
4.51 ppm to 4.32 ppm after the polymerization of TEFMA because
fluorinated methacrylate at d 4.49 ppm (monomer) and d 4.33 ppm
(polymer). For molecular weight evaluation, the withdrawn
sample was concentrated with nitrogen flow, then added hexane
to set out down all the polymer and then dried under vacuum for
48 h before 1H NMR analysis. The 1H NMR results for the evaluation
of molecular weight employed in the kinetic curve only consist of
samples more than 30% conversion based on after precipitation.
2.7. Chain extension using PTEFMA as macroinitiator
In
a representative run, acetonitrile (0.7 mL), HEMA-tms
(0.3011 g, 1.47 mmol) and bpy (11.5 mg, 0.0739 mmol) were
added in one flask of a two-connective flask set, while PTFEMA
(0.4172 g, or 0.0369 mmol. DP = 66 as evaluated from 1H NMR, and
Mw/Mn = 1.24), and CuCl (3.7 mg, 0.0369 mmol) were added to the
other flask. The reactants were bubbled for half an hour with argon
before subjected to three freeze–pump–thaw cycles, and then
transferred the liquid reactants to the flask containing solid
reactants. The brownish red solution was stirred and put into an
oil-bath preheated to 70 8C. The reaction was also stopped by
freezing with liquid nitrogen before exposure with air, and then,
precipitated out over ice. The precipitation was washed with
100 mL of hexane for 3 times before vacuum drying for 48 h,
generating 0.6842 g (monomer conversion = 89%) of product as
of the weakened p–p conjugation effect [56]. Also, the resonance
for –OCH3, normally at ꢀ3.7 ppm in poly(methyl methacrylate),
shifts to 4.31 ppm in PTFEMA due to the strong electron
withdrawing effect from –CF3 group in PTFEMA [57].
white powder. Mw/Mn = 1.27. 1H NMR (DMSO-d6):
–CH2CF3); 3.95 (br s, 2mH, –COOCH2– in the PHEMA-tms chains);
3.75 (br s, 2mH, –CH2OSi); 1.7–2.1 (br, 2(n + m)H, –CH2– in the
main chain); 0.8–1.3 (br, 3(n + m)H, –CH3 in the main chain); 0.1
d
4.33 (br s, 2nH,
The monofunctional initiator MEBrIB was extensively
employed for ATRP in our lab as a probe for 1H NMR analysis to
facilely assess the true molecular weight due to the resonance of
protons (–OCH2CH2OCH3) incorporated into the polymer chain
end located in a clear region of the spectrum [58]. However, this is
only possible for relatively low molecular weights (Mn < 15,000 g/
mol). Otherwise the signal of this group becomes negligible in
comparison with the signals of the other protons in the polymer
chain. Here, the average polymerization degree (DP) can be readily
calculated by comparing the peak intensity integration (S) of
3.35 ppm from –OCH3 (o) at the end of the polymer chain
introduced by initiator to that of 4.32 ppm from –CH2CF3 (m) at the
repeating unit of polymer chain with Eq. (1).
d
d
d
d
d
(br s, 9mH, –Si(CH3)3).
3. Results and discussion
3.1. ATRP of TFEMA using PMDETA as ligand
At the initial stage of this work, we roughly examined the ATRP
of TFEMA using MEBrIB as an initiator in the presence of CuBr as a
catalyst and PMDETA as a ligand in cyclohexanone at 70 8C. To
prepare objective homopolymer PTFEMA with theoretical maxim
repeating units of 400, the [M]0/[I]0/[L]0 was set at 400/1/1, while
solvent/momomer (S/M, v/v) was set at 1.3/1. Here [M]0, [I]0, and
3Sm
DP ¼
(1)
2So

566
G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
Fig. 3. 1H NMR spectra of (1) monomer, (2) initiator, and (3) PTFEMA recorded in CDCl3 at 30 8C.
The SEC results demonstrate that the number average
molecular weight and its polydispersity are summarized and listed
in Table 1 (entries 1–4).
molecular weight (Mn) of the resulting polymer is 100,450 g/
moL (DP = 598) and molecular polydispersity (Mw/Mn) is 1.52
using PS as standard, whereas the number average molecular
weight (Mn) of the resulting polymer as evaluated from 1H NMR
with Eq. (1) is 64,170 g/moL (DP = 382), the difference of Mn from
SEC to that from NMR is ascribed from difference of the
hydrodynamic volume of PTFEMA to that of PS used for calibrating
the SEC columns [52]. Moreover, the DP from gravimetric method
(monomer conversion ꢀ69%, here, the DP from conversion is based
on the assuming that an ideal living polymerization occurred) is
around 296, and that from 1H NMR is 382. This difference suggests
that irreversible termination, or low initiation efficiency, or loss of
active specie possibly occurred during the polymerization [22,23].
Careful inspection of the broad but monomodal SEC profile of this
sample prepared using PMDETA as a ligand reveals no obvious
evidence of permanent termination from recombination in the
initial and final stages of the polymerization, since no high
molecular weight tails and low molecular weight shoulders could
be observed (Fig. 4).
It clearly shows that the conversion increased with the increase
of the reaction time regardless of the ratio of [M]0/[I]0. For example,
a conversation of 56% was achieved after 4 h (entry 3 in Table 1)
compared to that of 81% (entry 3) after 6.5 h when [M]0/[I]0/[L] was
set to 135/1/1, while a conversation of 74% was obtained after 5.5 h
(entry 2) compared to that of 80% (entry 4) after 6 h when [M]0/[I]0/
[L] was set to 270/1/1. However, the molecular weight as evaluated
based on conversion from 1H NMR is lower than that from Eq. (1),
suggesting the possible loss of the active end group or irreversible
termination during the chain propagation. Moreover, the molecu-
lar weight polydispersity index (Mw/Mn) of the resulting polymer
samples is relatively higher, for example, ranging from 1.4 to 1.54,
especially when preparing the polymer with high targeting
polymerization degree and high conversion. This indicates that
TFEMA ATRP using PMDETA as the ligand was not well controlled.
To have better control on the preparation of PTFEMA, we thus
investigated the effect of other ligand, bpy, on ATRP of TFEMA in
terms of conversion, molecular weight and its polydispersity. The
results are also summarized and listed in Table 1 (entries 5–10).
The homopolymers with different molecular weights were
successfully synthesized with bpy as ligand judging by low
molecular weight polydispersity index (Mw/Mn, 1.11–1.29) for the
resulting polymer at high monomer conversion and high targeted
DP. In addition, a narrow PDI (Mw/Mn) was obtained when [CuBr]/
[bpy] was set to 1/2 compared to a broader one to that set to 1/1
(entries 6–9), which is consistent with the observation from others
when preparing polystyrene (PS) and poly(methyl acrylate) with
an optimum ligand-to-copper(I) halide ratio of 2/1 [56]. Moreover,
when a higher molecular weight was targeted by setting [M]0/[I]0/
[L] to 400/1/2 (entries 9 and 10), a relatively broader PDI and lower
monomer conversion were achieved compared to that with a lower
targeting molecular weight, which can be explained by the fact
that termination and other side reactions are also presented in
ATRP, and they become more prominent as higher molecular
weight polymers are targeted [60].
3.2. ATRP with various commonly used ligands
To investigate the details for ATRP of TFEMA with predesigned
average polymerization degree (DP) using PMDETA as ligand, we
then performed ATRP of TFEMA by varying [M]0/[I]0, polymeriza-
tion time, or S/M (v/v). The reaction recipe, monomer conversion,
As stated early, Me6TREN is commonly employed as represen-
tative tetradentate amine ligand in the ATRP of many monomers
[28]. In our system, as shown in entries 11–16 in Table 1, each
polymerization had a monomer conversion less than 30% when
Me6TREN wad used as a ligand. While Me6TREN was employed as a
ligand, the polymerization system was also opaque and heteroge-
neous. A very broad PDI higher than 2 was acquired at low
conversion. Moreover, the final product became gel, which is
insoluble in common solvents such as THF, acetone and DMF, at an
Fig. 4. SEC curve of PTFEMA prepared y using PMDETA as ligand.

G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
567
Table 1
Effect of ligand on ATRP of TFEMA in cyclohexanone.a
Convb (%)
DPc A/B
Mn (g/mol)
Mw/Mn
Tacticityf mm/mr/rr (%)
d
e
Entry
Ligand
[M]0/[I]0/[L]
[S]/[M] (v/v)
Time (h)
1
PMEDTA
PMEDTA
PMEDTA
PMEDTA
bpy
135/1/1
270/1/1
135/1/1
270/1/1
53/1/1
1.3
2
6.5
5.5
4
81
109/153
200/276
76/129
216/304
46/66
86/109
100/136
206/213
208/281
248/326
25/–
25,850
46,510
21,520
51,220
11,230
18,460
22,990
35,930
47,350
54,910
–
1.54
1.45
1.40
1.53
1.24
1.28
1.11
1.18
1.26
1.29
–
–/–/–
2
74
18/32/50
–/–/–
3
1.3
1.3
2
56
4
6
80
31/28/41
32/26/42
–/–/–
5
3
87
6
bpy
107/1/1
133/1/2
267/1/2
400/1/2
400/1/2
313/1/2
313/1/2
100/1/1
267/1/1
200/1/1
267/1/1
267/1/1
2
4.5
5
80
7
bpy
1.2
2
75
–/–/–
8
bpy
5
77
27/27/46
26/24/50
–/–/–
9
bpy
3
2.5
5
52
10
11
12
13
14
15
16
17
bpy
3
62
Me6TREN
Me6TREN
Me6TREN
Me6TREN
Me6TREN
Me6TREN
Me6TREN
2
3
8.0
21.3
20
11/37/52
–/–/–
3
5
67/1035
–
174,030
––
2.48
–
2
5
–/–/–
2
5
22
59/1005
–
168,980
–
3.22
–
–/–/–
2
7
Gel
Gel
Gel
21/30/49
–/–/–
2
10
10
–
–
–
2
–
–
–
–/–/–
a
The polymerization was carried out at 80 8C except for entries 1, 2, 3, 4, 5, which were all conducted at 70 8C.
Monomer conversion based on 1H NMR method.
b
c
d
e
f
A – calculated from MTFEMA ꢁ ConvNMR, and B – evaluated with Eq. (1) from 1H NMR.
Evaluated from 1H NMR.
SEC results using PS as standard.
Calculated from the method reported by Okamato et al. [59].
extending reaction time (as shown in entries 12–17). The activity
part. Fig. 5 shows the semilogarithmic kinetic plots for ATRP of
TFEMA with [M]0/[I]0/[Cu(I)Br]0/[L] of 100:1:1:2 for bpy system,
and 100:1:1:1 for PMDETA system. For bpy system, after a few
minutes, i.e. less than 6 min, the semilogarithmic kinetic plot is
linear (dashed line in Fig. 5). This demonstrates that the
polymerization rate is proportional to monomer concentration
and indicates that the concentration of growing radicals is constant
during the polymerization reactions after a few minutes [24,62].
It should be noted that the linear kinetic plot do not start in
origin, which should be discussed in detail later. Indeed, Hvilsted
group also found that the first-order plots for TFEMA ATRP using
ethyl 2-bromoisobutyrate as an initiator and N-(n-propyl)-2-
pyridylmethanimine (n-Pr-1) as a ligand at [M]0:[I]0:[CuBr]:[n-Pr-
1] = 117:1:1:2 do not start at origin for temperature above 80 8C,
i.e., 90 8C, 100 8C, and 110 8C [44]. We argue against the cause that
they claimed for this phenomena resulted from the time needed to
reach the required high temperature, since we found that kinetic
plot is very similar to that for the polymerization that was
performed with the reaction mixture heated at 80 8C for 10 min to
of N-based ligands in ATRP is correlated with the number of
coordinating sites as N4 > N3 > N2 ꢃ N1 in the heterogeneous
system [47]. In a heterogeneous system, Me6TREN/Cu(I) complex
would give more active propagating species with promoting
propagating rate, which would lead to the breaking of the balance
between the activation and deactivation in the ATRP process. In our
experiments, when conducting polymerization of TFEMA with
Me6TREN as a ligand, a low deactivator concentration and a very
fast activation would lead to an instant great increase in monomer
radical concentration, which leads to a higher apparent rate
constant (kapp) and high polydispersity. Generally, gel formation is
easily observed in systems where significant transfer to polymer
occurs, e.g., in acrylate polymerizations rather than methylacrylate
polymerization due to steric hinderance induced by a substituent
of the double bond. In the present system, high concentration of
propagating radical transfer to ethyl in the side group of the
polymer chain may possibly take responsibility for gel formation
during TFEMA ATRP using Me6TREN as a ligand.
To better understand the mechanism of ATRP, the stereochem-
istry of TFEMA polymerization was also investigated. The tacticity
of PTFEMA is calculated from 1H NMR of
a-methyl group according
to the signal assignments reported in the literature [59]. Indeed, as
shown in Table 1, PTFEMA homopolymers prepared using present
ATRP system have similar diad sequence composition to that
prepared by conventional radical polymerization using AIBN or
BPO initiator and within experimental error. This phenomenon has
already been found by several other groups, as demonstrated for
similar type of radical intermediate for ATRP as that for the
conventional free radical polymerization [61].
3.3. Kinetic study
To seek deep insight into the reasons for the effect of ligand on
ATRP of TFEMA, we further investigated the polymerization
kinetics from 1H NMR analysis in terms of the evolution of
monomer conversion by monitoring variation of sum of the peak
intensity for the double bond –CHdCH2 (5.85 and 6.19 ppm)
protons from TFEMA to the sum of that for –CH2–CF3 from
Fig. 5. Semilogarithmic kinetic plots for ATRP of TFEMA in cyclohexanone
(cyclohexanone/monomer (v/v) = 1/1) at 80 8C using ligand of PMDETA ( ), and
bpy ( ) with [M]0/[I]0/[Cu(I)Br]0/[L] = 100/1/1/X (X = 1 for PMDETA and X = 2 for
bpy), the dashed lines are only for the eyes. The five-star points represent the data
from bpy system initial addition of CuBr2 ([CuBr2]0/[CuBr]0 = 1/9). Inset is ln([M]0/
[M]) vs. t2/3 for PMDETA system (in which the dashed line represents best linear fit
for all the data with a correlation coefficient R2 = 0.95).
monomer at
d 4.49 ppm and polymer at d 4.33 ppm. Intermittent
sampling method was applied here to get samples for various
analyses, such as 1H NMR and SEC as detailed in the experimental

568
G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
Table 2
Polymerization phenomena of ATRP of TFEMA and solubility of CuBr (or CuBr2)/ligand complex.
Catalyst system
Polymerization phenomenaa
Solubility of ligand complexb
CuBr/ligand
CuBr2/ligand
[bpy]0/[CuBr]0 = 2/1
[PMDETA]0/[CuBr]0 = 1/1
The reaction mixture after mixed is brown and transparent. Green
precipitate presents after 10 min, and the whole reaction mixture
becomes green and turbid
1/2 brown
1/2 green precipitation
turbid (lightly soluble)
transparent (soluble)
The reaction mixture after mixed turns from transparent and light
green to cloudy and green right away, and then whole reaction
mixture successively becomes dark green and opaque
The reaction mixture after mixed is opaque, and the whole
reaction mixture gradually becomes transparent with little yellow
The reaction mixture after mixed is brown and opaque, green
precipitate presents after 6 min, and the whole reaction mixture
becomes green and turbid
1/1 Green transparent
(soluble)
1/1 blue semitransparent
(partially soluble)
[Me6TREN]0/[CuBr]0 = 1/1
1/1 colorless soluble
–
1/1 golden partially soluble
–
[bpy]0/[CuBr]0/[CuBr2]0 = 2/0.9/0.1
a
ATRP of TFEMA using different ligands in cyclohexanone (cyclohexanone/monomer, v/v = 1/1) at 80 8C.
CuBr (or CuBr2)/ligand complex is dispersed in cyclohexanone/monomer (v/v = 1) with [monomer]/[CuBr (or CuBr2)] = 100/1, the solubility and color are observed in a
b
sealed flask free of oxygen.
reach equilibrium before the addition of MEBrIB to initiate the
polymerization (defined t = 0 when MEBrIB was added).
However, for PMDETA system as shown in Fig. 5, the curved
plot demonstrates that the radical concentration is continuously
decreased during the polymerization reactions because of contin-
uous consumption of monomer, irreversible termination of
radical–radical coupling or a side reaction involving a radical
transfer from the propagating chain to the ligand species to form
nonpropagatable (dead) chains, as has been proposed by other
research groups [63,64].
There were two main kinetic descriptions of the conversion vs.
time, one is proposed by Matyjaszewski et al. [48,57], which is
based on a constant radical concentration due to the equilibrium
between polymer radicals, dormant chains and copper species (Eq.
(2), denoted as M-equ.), while the other one developed by Fischer
et al., [65,66] is based on the persistent radical effect (Eq. (3),
denoted as F-equ.)
the reactivity of the monomer compared with m. However, the
conclusions are quite different from those of different research
groups. For example, Ameduri et al. reported that polymerizations
of fluorinated (meth)acrylates proceed rather slowly compared to
nonfluorinated monomers [69], on the contrary, Beuermann et al.
declared that the propagation rate coefficient (kp) for the
fluorinated monomer is higher than that of corresponding
nonfluorinated methacrylate monomer at identical temperature,
possibly because of less interactions between the macroradicals
compared to nonfluorinated systems [70]. Our finding is consistent
with the results from Beuermann et al., since the apparent rate
constants of MMA ATRP have previously been reported in the range
of (0.10–5) ꢁ 10ꢂ4 sꢂ1, which is lower than that from present study
[71].
As stated before, the activity of CuII and CuI/ligand depends
dramatically on their solubility in the polymerization medium
[24]. It is interesting to ask why the ATRP of TFEMA in these three
ligand systems is heterogeneous but exhibits obvious controlla-
bility. To answer this, further evidence for the solubility of CuBr (or
CuBr2)/ligand complex is included in Table 2. It demonstrates that
the polymerization was heterogeneous because CuBr2/ligand (see
mechanism illustrated in Fig. 6) was not very soluble in the
polymerization media. In these systems, the deactivator dissolves
poorly or precipitates out of the system during the reaction and as a
consequence will show a ceiling CuII concentration [72,73].
For bpy system, possibly because of the insolubilization rate
½Mꢄ0
½Mꢄ
½RXꢄ0½CuIꢄ
½CuIIꢄ0
ln
ln
¼ kpKeq
0 t ¼ KMatyjasszewski
t
(2)
(3)
ꢀ
ꢁ
1=3
1=3
½Mꢄ0
½Mꢄ
3
Keq
¼
kpð½RXꢄ0½CuIꢄ0Þ
t2=3 ¼ KFischer
t
2
3kt
As the kinetic plots shown in Fig. 5, Eq. (2) can apply in bpy
system, indicating that, after a few minutes, the atom transfer
equilibrium constant (Keq = ka/kd) and/or the CuI and/or initiator
concentrations are sufficiently low and the CuII concentration is
sufficiently high [67,68]. However, inset in Fig. 5 is ln([M]0/[M]) vs.
coefficients (kinsol) are higher than the solubilization rate
coefficients (ksol, Fig. 6), an equilibrium concentration of copper
species will be reached only after a few minutes, judging by the
color change during the polymerization after mixing all the
reactants, and by similar polymerization phenomena for the bpy
system with initial addition of CuBr2 ([CuBr2]0/[CuBr]0 = 1/9) to
that with only CuBr (Table 2). This is also demonstrated by the fact
that the kinetic plot is almost the same for bpy system without
addition of CuBr2 to that for bpy system with initial addition of
CuBr2([CuBr2]0/[CuBr]0 = 1/9). Two conclusions might be drawn.
Firstly, the initial CuII concentration before the addition of initiator
t
2/3 for PMDETA system, the plot is linear suggesting that the data is
well fitted with Eq. (3), which means that after a very short time
[CuII] ꢃ [R*]. In this case the radical concentration is controlled by
the concentration of the activator, deactivator and ‘dormant’
chains and from that point onwards all reactant concentrations can
be effectively calculated in time, and during the polymerization the
concentrations of both RX and CuI are continuously decreasing,
resulting in a decrease of the polymerization rate [65,66].
The slopes of F-equ and M-equ kinetic plots yielded the
apparent rate coefficients of 15.2 ꢁ 10ꢂ4 sꢂ1 and 2.7 ꢁ 10ꢂ4 sꢂ1 for
PMDETA and bpy systems, respectively. These are in reasonable
agreement with the values of 1.6 ꢁ 10ꢂ4 sꢂ1 to 2.9 ꢁ 10ꢂ4 sꢂ1
reported for ATRP of TFEMA using ethyl 2-bromoisobutyrate as
initiator and N-(n-propyl)-2-pyridylmethanimine (n-Pr-1) as
ligand at [M]0:[I]0:[CuBr]:[n-Pr-1] = 117:1:1:2 at 80–110 8C [44].
So far, there have been a few reports on the kinetics of
fluorinated (meth)acrylates in free radical polymerizations as well
as in RAFT or ATRP. For monomers of type CH255CHCOO–(CH2)n–
CmF2m+1, it was demonstrated that n has a significant influence on
Fig. 6. Schematic illustration for heterogeneous ATRP mechanism of TFEMA.

G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
569
supporting a living process of TFEMA polymerization initiated
with bpy system [52].
The polydispersities for all the PTFEMA obtained by ATRP using
bpy as ligand are relatively low and decrease with conversion and
increasing chain length, i.e., reaching value of ca. 1.1 at about 80%
conversion. At the early stage of the reaction, the polymerization
system had a higher initial CuI concentration and low amount of
dissolved CuII/ligand complex due to low solubilization coefficients
(ksol). This resulted in a high radical flux and therefore a high
termination rate. Irreversible radical termination can cause a
decrease in initiation efficiency and high PDI. After the initial
period the reaction rate decreases and so does the termination rate,
and at higher conversions the polydispersity indices decrease [75].
However, for PMDETA system, it shows that although the Mn
increases linearly with conversion, the measured values of Mn
deviates from the theoretical prediction significantly under the
reaction conditions used here. This may be due to slow exchange
reactions which result in low initiation efficiencies and the
termination by radical-radical coupling. The polydispersity indices
Mw/Mn were relatively broad (more than 1.40) that may be because
of a gradual loss of reactive species resulting from irreversible
termination which can be seen in the first-order plot. Moreover, on
the contrary to that from bpy system, the polydispersities for all
the PTFEMA increase with conversion and increasing chain length,
suggesting that that the chain transfer may become increasingly
significant with the increase of molecular weights for PMDETA
system [75]. All these further demonstrate that the PMDETA-based
system provides the undesirable characteristics of low initiation
efficiency, high polydispersity, and uncontrollable apparent
polymerization rate, and thus bad control over polymerization
process.
Fig. 7. The dependence of molecular weight, Mn (as evaluated from 1H NMR), and
molecular weight polydispersity (Mw/Mn) upon monomer conversion for ATRP of
TFEMA at 80 8C with various ligands of PMDETA ( ), and bpy (&) (the straight line
represents the theoretical molecular weight at certain monomer conversion, the
dotted curves are only for the eyes).
is lower than the equilibrium value. Although the CuI concentra-
tion is relatively high, owing to the higher oxidation potential of
the bpy complexed metal center than that of PMDETA or Me6TREN
complexed metal center, leading to a greater tendency toward
radical deactivation and lower overall concentration of radical
species. A lower irressible termination is noted in the initial stage
of the polymerization [28–30]. In other words, after a few minutes,
the dissolved CuII is constant and high enough to minimize
termination events, as a result a controlled polymerization with a
low fraction of dead polymer chains is expected. Secondly, the
initially added solid CuBr2 does not dissolve well [74] and as a
result will not contribute to the control of the polymerization
leading to a similar kinetic plot to that without CuBr2. However, for
PMDETA or Me6TREN system, although the solubilization coeffi-
cients (ksol) are high enough judging by the fast change of color
after mixing all the reactants, the deactivation rate of growing
TFEMA chains is not fast enough in comparison with PFEMA
propagation. These are due to the electronic properties of these
two ligands. This combination results in bad or even loss of control
on TFEMA ATRP.
Fig. 7 shows the dependence of number average molecular,
Mn,NMR (as evaluated from 1H NMR), and polydispersity (Mw/Mn)
vs. conversion for ATRP of TFEMA with PMDETA and bpy as ligand,
the straight line in the plot represents the theoretical molecular
weight at certain monomer conversion. For bpy system, the Mn
evolves linearly with conversion, and measured values of Mn are
very close to the theoretical prediction based upon ([M]0 ꢂ [M]t)/
[In]0 under the reaction conditions used here. Moreover, an
initiator efficiency derived from Mn,NMR/Mtheoretical is from 68% for
30% of monomer conversion to 95% for 88% of monomer
conversion, indicating slow initiating but a high initiator efficiency,
and thus small contribution of irreversible transfer, further
3.4. Chain extension
To further figure out the effect of ligand, PTFEMA-Br derived
from ligand bpy or PMDETA was employed as macroinitiator for
block copolymerization with HEMA-tms, to confirm the ‘‘living’’
chain end of the polymer (Fig. 8).
Table 3 lists the data for the chains extension reaction using the
two macroinitiators, PTFEMA-Br with similar molecular weight
(Mn) derived from bpy system and PMDETA system, respectively.
The chain extension reactions were both carried out in acetonitrile
at 70 8C with [HEMA-tms]0/[PTFEMA-Br]0/[CuCl]0/[bpy]0 = 40/1/1/
2.
The conversion of HEMA-tms reached about 89% as evaluated
from 1H NMR after 23 h using PTFEMA-Br as a macroinitiator from
bpy system, indicating a high initiating efficiency of the initiator.
The SEC curve of the resultant block copolymer shifts toward high
molecular weight, and is symmetric and unimodal (Fig. 9a). The
successful extension of the well-defined PHEMA-tms block further
demonstrates ‘‘living’’ character of the macroinitiator and well-
controlled ATRP of TFEMA using bpy as ligand. On the contrary, the
conversion of HEMA-tms came to only about 41% after 23 h using
Fig. 8. Chain extension from ATRP using PTFEMA-Br as macroinitiator.

570
G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
Table 3
Chain extension using PTFEMA-Br as microinitiator.a
Entry
Macroinitiator
Block copolymer
b
c
c
Mn (g/mol)
Mw/Mn
Monomer conversionb (%)
Mw/Mn
1
2
11,100
10,800
1.26
1.62
89
41
1.27
2.12
a
Reaction condition: [HEMA-tms]0/[PTFEMA-Br]0/[CuCl]0/[bpy]0 = 40/1/1/2 at 70 8C, reaction time = 23 h, the macroinitiator PTFEMA-Br in entries 1 and 2 was prepared
using bpy and PMDETA as ligand, respectively.
b
Evaluated from 1H NMR.
c
SEC performed using THF as eluent and narrow dispersed polystyrene as standard.
Fig. 9. SEC curves of (a) macroinitiator derived from ATRP of TEFMA using bpy as ligand and PTFEMA-b-PHEMA-tms from the macroinitiator, and (b) maroinitiator derived
from ATRP of TEFMA using PMDETA as ligand and PTFEMA-b-PHEMA-tms from the macroinitiator.
Table 4
Bulk properties of the prepared polymer.
Entry
Mn (NMR) (g/mol)
Mw/Mn (SEC)
Tg (DSC) (8C)
Thermo-stability (TGA)
Static contact angle (8)
TSWL (8C)
T10WL (8C)
WL300 (%)
1
2
3
4
11,090
25,370
38,800
51,240
1.21
1.11
1.18
1.29
48
65
74
75
108
131
162
168
213
213
238
246
69
64
37
29
82 ꢅ 2
85 ꢅ 2
98 ꢅ 3
96 ꢅ 2
PTFEMA-Br as macroinitiator from PMDETA system. The low
initiating efficiency for PTFAMA-Br derived from PMDETA system
further suggests that the loss of Br and chain transfer occurred
during the polymerization of TFEMA [52,76]. Fig. 9b shows SEC
curves of macroinitiator derived from ATRP of TFEMA using
PMDETA as ligand and PTFEMA-b-PHEMA-tms from this macro-
initiator. The SEC curve of the resultant block copolymer is not only
board but also asymmetric, further demonstrating bad-controlled
ATRP of TFEMA using PMDETA as ligand.
that a Tg of 74 8C with Mn = 384,000 g/mol and 78 8C with
Mn = 574,000 g/mol for PTFEMA sample was due to differences
of their syndiotacticity [44]. In fact, when Tg is plotted against 1/Mn
(Fig. 10), an almost linear relationship is demonstrated (R2 = 0.98),
a maximum Tg of 82 8C is reasonably extrapolated from the fitted
line.
3.5. Bulk properties of the homopolymer PTFEMA
We also examined the bulk properties such as, glass transition
temperature, thermo-stability, and film hyrophobicity of the
PTFEMA homopolymer with different molecular weights. Table
3 lists the data of PTFEMA with variation of number average
molecular weight (Mn) Table 4.
The Tg of PTFEMA increases from 48 8C for the sample with
Mn = 11,090 g/mol to about 75 8C for the sample with
Mn = 51,240 g/mol. This is believed to be because of the fact that
the mobility of chain segment decreased for samples with higher
Mn. Above a critical Mn, Tg should no longer increase with the
molecular weight. [59] It was reported that the PTFEMA with
Mn = 8600 g/mol exhibits a Tg at 59 8C. [59] It was also reported
Fig. 10. The dependence of glass transition temperature of PTFEMA on the number
average molecular weight (Mn).

G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
571
4. Conclusion
In summary, we have demonstrated that well-defined low-
fluorinated homopolymer poly(trifluoroethyl methacrylate),
PTFEMA, could be prepared by heterogeneous atom transfer
radical polymerization (ATRP) using didentate amines, bpy as
ligand in cyclohexanone in the presence of initiator methoxyl
ethylene 2-bromoisobutyrate. The obtained PTFEMA has
number-average molecular weight (Mn) close to the calculated
value and relatively narrow polydispersity. The resulting
a
a
homopolymer with high molecular weight shows higher
thermo-stability.
Acknowledgements
Fig. 11. TGA curves for PTFEMA samples with various molecular weights, the
We thank the Project of Leading Talent of Guangdong Province,
the Outstanding Overseas Chinese Scholars Funds of the Chinese
Academy of Sciences and NSF of China for financial support of this
project. Dr. Ian Wyman is thanked for language polishing.
numbers 1–4 correspond to that in Table 4.
References
Fig. 11 shows thermo-stability profiles of the samples with
different molecular weights as evaluated from the thermo-
gravimetric analysis (TGA), and some of the data are included in
Table 3. A two-stage degradation behavior with two distinctive
plateaus for all the samples is observed which is in good agreement
with the observation by other researchers [41,44]. The starting
temperature for weight loss (TSWL) significantly increases with the
increase of the molecular weight (Mn), for example, TSWL = 108 8C
increases to 168 8C as the Mn = 11,090 g/mol increases to 51,240 g/
mol, moreover, the temperature for 10% weight loss (T10%WL) is at
213 8C, 213 8C, 238 8C, and 246 8C for PTFEMA sample with
Mn = 11,090, 25,370, 38,800, and 51,240 g/mol, respectively. These
indicate that the sample with higher molecular weight is more
stable compared to that with lower molecular weight. The weight
loss at 300 8C (WL300), which is almost corresponding to the
temperature for the beginning of the second plateau in the TGA
profiles of all the samples (Fig. 9), decreases ranging from 69% to
28% as the molecular weight increases from 11,090 to 51,240 g/
mol, since the content for –CF3, –CH2CF3, –OCH2CF3, and –
COOCH2CF3 in the polymer chain is around 40, 49, 58, and 75%,
respectively. This suggests that the degradation behavior remark-
ably depends on the molecular weight of PTFEMA, for example, for
sample with low molecular weight (Mn = 11,090, and 25,370 g/
mol), the weight loss routes for first plateau might be ascribed from
bond cleavage either in the ester group (between the carbonyl and
oxygen), between the ester group and the fluorinated ethyl, or in
the ethyl pendant chain [44], and degradation of main chain, while
the second plateau is mainly from the degradation of main chain,
for sample with high molecular weight (Mn = 38,800, and 51,240 g/
mol), however, the weight loss (less than 37%) for first plateau
might be mainly ascribed from the loss of –CF3, and the second
plateau derives from the cleavage of part of the side chain and
degradation of main chain. The detailed thermo-degradation
mechanism depends on further study, such as FTIR–TG and TG–
MS and will be reported in another paper.
[1] N.M.L. Hansen, K. Jankova, S. Hvilsted, Eur. Polym. J. 43 (2007) 255–293.
[2] R.J. Plunkett, U.S. Patent 2, 230, 654 (1961).
[3] J. Scheirs, Modern Fluoropolymers, Wiley, New York, 1997.
[4] I.J. Park, S.B. Lee, C.K. Choi, J. Appl. Polym. Sci. 54 (1994) 1449–1454.
[5] R.D. van de Grampel, Surfaces of Fluorinated Polymer Systems, Technische
Universiteit Eindhoven, 2002.
[6] G. Hougham, K. Johns, P.E. Cassidy, Fluoropolymers: Synthesis and Properties,
Plenum, New York, 1999.
[7] M. Morita, H. Ogisu, M. Kubo, J. Appl. Polym. Sci. 73 (1999) 1741–1749.
[8] J.G. Riess, Curr. Opin. Colloid Interface Sci. 14 (2009) 294–304.
[9] J. Eastoe, S. Gold, D.C. Steytler, Langmuir 22 (2006) 9832–9842.
[10] Z. Zeng, B.S. Phillips, J.C. Xiao, J.M. Shreeve, Chem. Mater. 20 (2008) 2719–2726.
[11] C. Huang, H.E. Katz, J.E. West, Langmuir 23 (2007) 13223–13231.
[12] E.M.W. Tsang, Z.B. Zhang, A.C.C. Yang, Z.Q. Shi, T.J. Peckham, R. Narimani, B.J.
Frisken, S. Holdcroft, Macromolecules 42 (2009) 9467–9480.
[13] J.E. Hensley, J.D. Way, Chem. Mater. 19 (2007) 4576–4584.
[14] L.M. Tang, Y. Li, X.M. Wu, X.F. Shan, W.C. Wang, Adv. Technol. 15 (2004) 39–42.
[15] T. Imae, Curr. Opin. Colloid Interface Sci. 8 (2003) 307–314.
[16] S.J. McLain, B.B. Sauer, L.E. Firment, Macromolecules 29 (1996) 8211–8219.
[17] K. Tadano, Y. Tanak, T. Shimizu, S. Yano, Macromolecules 32 (1999) 1651–1660.
[18] M. Eberhardt, R. Mruk, R. Zentel, P. The´ato, Eur. Polym. J. 41 (2005) 1569–1575.
[19] (a) O.W. Webster, Science 251 (1991) 887;
(b) Q.H. Zhang, X.L. Zhan, F.Q. Che, J. Appl. Polym. Sci. 104 (2007) 641–647.
[20] G. Moad, E. Rizzardo, S.H. Thang, Acc. Chem. Res. 41 (2008) 1133–1142.
[21] P.B. Zetterlund, Y. Kagawa, M. Okubo, Chem. Rev. 108 (2008) 3747–3794.
[22] (a) B. Ameduri, B. Boutevin, Well Architectured Fluoropolymers: Synthesis Prop-
erties and Applications, Elsevier, Amsterdam, 2004;
(b) P. Lacroix-Desmazes, B. Ameduri, B. Boutevin, Collect. Czech. Chem. Commun.
67 (2002) 1383–2141;
(c) B. Ameduri, Macromolecules 43 (2010) 10163–10184.
[23] (a) D. Yang, L. Tong, Y.J. Li, J.H. Hu, S. Zhang, X.Y. Huang, Polymer 51 (2010) 1752–
1760;
(b) Y.j. Li, S. Zhang, H. Liu, Q.N. Li, W.X. Li, X.Y. Huang, J. Polym. Sci. Polym. Chem.
48 (2010) 5419–5429;
(c) L. Tong, Z. Shen, D. Yang, S. Chen, Y.J. Li, J.H. Hu, G.L. Lu, X.Y. Huang, Polymer 50
(2009) 2341–2352;
(d) Y.J. Li, S. Chen, S. Zhang, Q.N. Li, G.L. Lu, W.X. Li, H. Liu, X.Y. Huang, Polymer 50
(2009) 5192–5199.
[24] (a) J.H. Xia, K. Matyjaszewski, Chem. Rev. 101 (2001) 2921–2990;
(b) N.V. Tsarevsky, K. Matyjaszewski, Chem. Rev. 107 (2007) 2270–2299;
(c) A.P. Narrainen, S. Pascual, D.M. Haddleton, J. Polym. Sci. Part A: Polym. Chem.
40 (2002) 439–450.
[25] (a) M. Kato, M. Kamigaito, M. Sawamoto, T. Higashimura, Macromolecules 28
(1995) 1721–1723;
Homopolymer solutions in THF at 0.2 mg/mL with different Mn
were dropped on a cleaned slide glass and dried at room
temperature in a desiccator before water contact angle analysis.
As shown in Table 3, the static contact angle of water on PTFEMA
films are ranged in 82–988. Moreover, it seems that film from lower
molecular weight sample has lower contact angle of water than
that from higher molecular weight sample, for example, the
contact angle for film from sample with Mn = 11,090 g/mol is 828,
whereas that for film from sample with Mn = 51,240 g/mol is 968
(entries 1 and 4 in Table 3), which might be because of pronounced
‘‘end-group effect’’ for the sample with low molecular weight.
(b) J.S. Wang, K. Matyjaszewski, J. Am. Chem. Soc. 117 (1995) 5614–5615.
[26] (a) M.H. Acar, N. Bicak, J. Polym. Sci. Part A: Polym. Chem. 41 (2003) 1677–1680;
(b) A. Mittal, S. Sivaram, J. Polym. Sci. Part A: Polym. Chem. 43 (2005) 4996–5008.
[27] J. Xia, X. Zhang, K. Matyjaszewski, ACS Symposium Series 760, American Chemical
Society, Washington, DC, 2000, p. 207.
[28] J. Qiu, K. Matyjaszewski, L. Thouin, C. Amatore, Macromol. Chem. Phys. 2000 (201)
(2000) 1625–1631.
[29] R. Venkatesh, S. Harrisson, D.M. Haddleton, B. Klumperman, Macromolecules 37
(2004) 4406–4416.
[30] R.K. O’Reilly, V.C. Gibson, A.J.P. White, D. Williams, J. Polyhedron 23 (2004) 2921–
2928.
[31] W. Tang, N.V. Tsarevsky, K. Matyjaszewski, J. Am. Chem. Soc. 128 (2006) 1598–
1604.
[32] A. Hirao, K. Sugiyama, H. Yokoyama, Prog. Polym. Sci. 32 (2007) 1393–1438.

572
G. He et al. / Journal of Fluorine Chemistry 132 (2011) 562–572
[33] (a) A. Tuteja, W. Choi, M.L. Ma, J.M. Mabry, S.A. Mazzella, G.C. Rutledge, G.H.
McKinley, R.E. Cohen, Science 318 (2007) 1618–1622;
(b) M. Lazzari, M. Aglietto, V. Castelvetro, O. Chiantore, Polym. Degrad. Stabil. 79
(2003) 345–351.
[34] S.A. Brewer, C.R. Willis, Appl. Surf. Sci. 254 (2008) 6450–6454.
[35] Z.J. Wei, W.L. Liu, D. Tian, C.L. Xiao, X.Q. Wang, Appl. Surf. Sci. 2010 (256) (2010)
3972–3976.
[53] J.W. Hu, G.J. Liu, G. Nijkang, J. Am. Chem. Soc. 130 (2008) 3236–3237.
[54] T. Kakiuchi, M. Nakanishi, M. Senda, Bull. Chem. Soc. Jpn. 61 (1988) 1845–1851.
[55] A.E. Acar, M.B. Yagci, L.J. Mathlas, Macromolecules 33 (2000) 7700–7706.
[56] Z.T. Liu, J.G. Chen, Z.W. Liu, J. Lu, Macromolecules 41 (2008) 6987–6992.
[57] J.L. Wang, T. Grimaud, K. Matyjaszewski, Macromolecules 30 (1997) 6507–6512.
[58] F.B. Zhao, Z.L. Liu, L. Feng, J.P. Sun, J.W. Hu, J. Polym. Sci. Part B: Polym. Phys. 47
(2009) 1345–1355.
[36] J.G. Riess, M.P. Krafft, in: R.M. Winslow (Ed.), Blood substitutes, Elsevier, 2006
(Chapter 24).
[59] W.H. Liu, K. Tang, Y.Z. Guo, Y. Koike, Y. Okamoto, J. Fluorine Chem. 123 (2003)
147–151.
[37] M.G. Dhara, S. Banerjee, Prog. Polym. Sci. 35 (2010) 1022–1077.
[38] S.D. Personick, Fiber Optic Technology and Applications, Plenum Press, New York,
1985 (Chapter 2).
[39] L. Yi, X. Zhan, F. Chen, L. Huang, F. Du, J. Polym. Sci. Part A: Polym. Chem. 43 (2005)
4431–4438.
[40] J. Roussel, B. Boutevin, J. Fluorine Chem. 108 (2001) 37–45.
[41] X.Y. Liu, J. Chen, P. Sun, Z.W. Liu, Z.T. Liu, React. Funct. Polym. 70 (2010) 972–979.
[42] T.Y. Guo, D. Tang, M. Song, B. Zhang, J. Polym. Sci. Part A: Polym. Chem. 45 (2007)
5067–5075.
[43] S. Perrier, S.G. Jackson, D.M. Haddleton, B. Ameduri, B. Boutevin, Macromolecules
36 (2003) 9042–9049.
[60] K. Matyjaszewski, K. Davis, T. Patten, M. Wei, Tetrahedron 53 (1997) 15321–
15329.
[61] D.M. Haddleton, C.B. Jasieczek, M.J. Hannon, A.J. Shooter, Macromolecules 30
(1997) 2190–2193.
[62] K.A. Davis, H. Paik, K. Matyjaszewski, Macromolecules 32 (1999) 1767–1776.
[63] M. Bednarek, T. Biedron, P. Kubisa, Macromol. Rapid Commun. 20 (1999) 59–65.
[64] R. Sharma, A. Goyal, J.M. Caruthers, Y.Y. Won, Macromolecules 39 (2006) 4680–
4689.
[65] H. Fischer, Macromolecules 30 (1997) 5666–5672.
[66] H. Fischer, J. Polym. Sci. Part A: Polym. Chem. 37 (1999) 1885–1901.
[67] T. Fukuda, A. Goto, K. Ohno, Macromol. Rapid Commun. 21 (2000) 151–165.
[68] H. Zhang, B. Klumperman, W. Ming, H. Fischer, R. van der Linde, Macromolecules
34 (2001) 6169–6173.
[44] N.M.L. Hansen, M. Gerstenberg, D.M. Haddleton, S. Hvilsted, J. Polym. Sci. Part A:
Polym. Chem. 46 (2008) 8097–8111.
[45] L.F. Zhang, Z.P. Cheng, N.C. Zhou, R.M. Zhang, X.L. Zhu, Macromol. Symp. 261
(2008) 54–63.
[69] B. Guyot, B. Ameduri, B. Boutevin, M. Melas, M. Viguier, A. Coller, Macromol.
Chem. Phys. (199) (1998) 1879–1885.
[46] H. Peng, I. Blakey, B. Dargaville, F. Rasoul, S. Rose, A.K. Whittaker, Biomacromo-
lecules 10 (2009) 374–381.
[47] W. Du, A.M. Nystro¨m, Z. Lei, K.T. Powell, Y. Li, C. Cheng, S.A. Wickline, K.L. Wooley,
Biomacromolecules 9 (2008) 2826–2833.
[48] J.S. Wang, K. Matyjaszewski, Macromolecules 28 (1995) 7901–7910.
[49] X.S. Wang, N. Luo, S.K. Ying, J. Polym. Sci. Part A: Polym. Chem. 37 (1999) 1255–
1263.
[50] J. Xia, K. Matyjaszewski, Macromolecules 30 (1997) 7697–7700.
[51] R. Kroll, C. Eschbaumer, U.S. Schubert, M.R. Buchmeiser, K. Wurst, Macromol.
Chem. Phys. 202 (2001) 645–653.
[70] R. Siegmann, S. Beuermann, Macromolecules (43) (2010) 3699–3704.
[71] M.H. Acar, C.R. Becer, H.A. Ondur, S. Inceoglu, Polym. Preprints 46 (2005) 433–
434.
[72] D.A. Shipp, K. Matyjaszewski, Macromolecules 33 (2000) 1553–1559.
[73] M.A.J. Schellekens, F. de Wit, B. Klumperman, Macromolecules 34 (2001) 7961–
7966.
[74] M. van der Sluis, B. Barboiu, N. Pesa, V. Percec, Macromolecules 31 (1998) 9409–
9412.
[75] K. Matyjaszewski, J. Phys. Org. Chem. 8 (1995) 197.
[76] M. Bednarek, K. Jankova, S. Hvilsted, J. Polym. Sci. Part A: Polym. Chem. 45 (2007)
333–334.
[52] L. Feng, J.W. Hu, Z.L. Liu, F.B. Zhao, Polymer 48 (2007) 3616–3623.