696
LITHOGRAPHIC RESISTS
Vol. 6
LIVING RADICAL POLYMERIZATION
Introduction
Conventional free-radical polymerization (FRP) is characterized by the continuous
formation of radicals (initiation) and continuous disappearance of radicals (termi-
nation) (see R
ADICAL POLYMERIZATION
). The statistical nature of the FRP process
leads to relatively polydisperse polymers, with a typical instantaneous polydisper-
sity index (PDI) of 1.5 or 2, depending on the mode of termination (1). The complete
growth of an individual chain in FRP takes typically in the order of seconds. The
molar mass distribution (MMD) of an FRP product consists of an accumulation of
instantaneous MMDs formed over the course of the reaction. Hence the final PDI
may have a value significantly larger than 1.5 or 2. FRP is a versatile technique
that allows the polymerization of monomers with various functional groups. Fur-
thermore, the polymerization can be carried out in a variety of reaction media,
ranging from bulk monomer, via solution polymerization, to heterogeneous media,
eg emulsion and suspension polymerization, and polymerization in supercritical
carbon dioxide. One of the main disadvantages of FRP is the poor control over the
microstructure of the synthesized macromolecules. This includes the relatively
high PDI, and also the practical impossibility to synthesize block copolymers, and
other advanced structures.
Advanced structures can be synthesized via living polymerization tech-
niques. Notable example of these techniques is anionic polymerization (2), which is
known to allow the synthesis of low PDI materials as well as block copolymers. The
Encyclopedia of Polymer Science and Technology. Copyright John Wiley & Sons, Inc. All rights reserved.
Vol. 6
LIVING RADICAL POLYMERIZATION
697
main disadvantages of anionic polymerization are the limited choice of monomers,
and the extremely demanding reaction conditions.
In order to overcome the disadvantages of FRP without sacrificing the above-
mentioned advantages, it was recognized that a living character had to be realized
in conjunction with the free-radical mechanism. At present three main mecha-
nisms exist that ensure this living character by establishing an equilibrium be-
tween active (radical) and dormant chains. In a typical living radical polymeriza-
tion (LRP), all chains are started early in the reaction, and are allowed to grow
throughout the reaction. How this is achieved for individual LRPs will be detailed
below. In general, the result of a successful LRP will be a polymer with low PDI,
and predetermined (number-average) molar mass.
In this overview, the historical development of techniques will be summa-
rized. In addition, the mechanisms and kinetics of the main three techniques
will be highlighted. Finally, scope and limitations of the three techniques will be
discussed on the basis of a number of examples.
Historical Overview
Probably the first publication that dealt with LRP already appeared in the late
1960s (3). In the 1970s and 1980s it was realized that the main drawback in FRP,
ie, the broad polydispersities, and the inability to form block copolymers could
be overcome by the introduction of reversible deactivation of the active radical.
Otsu reported on several attempts to introduce this reversibility (4,5). The inifer-
ter technique was developed, which comprises a UV-labile or thermally labile link
that can be reversibly cleaved, and used for monomer insertion (6,7). Otsu showed
that the technique allowed the synthesis of block copolymers, albeit with a PDI
around 2. Currently it turns out that the iniferter technique may still have ad-
vantages over the more recent developments, eg when the synthesis of multiblock
copolymers is aimed for (8).
In the mid-1980s, the first technique that relies on the reversible termination
of radicals with a stable free radical was developed in the group of E. Rizzardo
at CSIRO in Australia. Rizzardo and co-workers found that nitroxide-stable free
radicals were able to add to carbon-centered radicals to form alkoxy amines (9).
In certain cases these alkoxy amines are thermally unstable, so that they enter
into an equilibrium between (transient) carbon-centered radical and (persistent)
nitroxide radical on one side, and alkoxy amine on the other side. TEMPO was ini-
tially the most frequently used nitroxide in conjunction with the polymerization
of styrene and its derivatives. The TEMPO–polystyrene adduct requires temper-
atures of 120
◦
C or above in order to establish an equilibrium at which polymer-
ization takes place. Around the mid-1990s Georges and co-workers focused on the
TEMPO-mediated polymerization of styrene (10), and developed various strate-
gies to overcome intrinsic weaknesses of the system. They used camphor sulfonic
acid to enhance the rate of polymerization (11). This rate enhancement was later
elucidated to be due to the destruction of excess nitroxide that builds up during
the polymerization.
By the end of the 1990s, Hawker and Tordo independently developed alter-
native nitroxides that allowed polymerization at temperatures below 100
◦
C, and
also allowed the polymerization of acrylates (12,13).
698
LIVING RADICAL POLYMERIZATION
Vol. 6
A parallel development was initiated by the first publications from Sawamoto
and Matyjaszweski. They reported independently on the transition-metal–
catalyzed polymerization of various vinyl monomers (14,15). The technique, which
was termed atom transfer radical polymerization (ATRP), uses an activated alkyl
halide as initiator, and a transition-metal complex in its lower oxidation state as
the catalyst. Similar to the nitroxide-mediated polymerization, ATRP is based on
the reversible termination of growing radicals. ATRP was developed as an exten-
sion of atom transfer radical addition (ATRA), the so-called Kharasch reaction (16).
ATRP turned out to be a versatile technique for the controlled polymerization of
styrene derivatives, acrylates, methacrylates, etc. Because of the use of activated
alkyl halides as initiators, the introduction of functional endgroups in the polymer
chain turned out to be easy (17–22). Although many different transition metals
have been used in ATRP, by far the most frequently used metal is copper. Nitrogen-
based ligands, eg substituted bipyridines (14), alkyl pyridinimine (Schiff ’s base)
(23), and multidentate tertiary alkyl amines (24), are used to solubilize the metal
salt and to adjust its redox potential in order to match the requirements for an
ATRP catalyst. In conjunction with copper, the most powerful ligand at present is
probably tris[2-(dimethylamino)ethyl)]amine (Me
6
-TREN) (25).
Many different polymer architectures have been synthesized via ATRP.
These architectures include block copolymers (14), graft copolymers (26–29), star-
branched polymers, etc. Some work was devoted to the reduction of the amount
of catalyst used, or to the efficient removal of catalyst from the resulting polymer
(30–33).
The third large development is, unlike nitroxide-mediated polymerization
and ATRP, based on reversible chain transfer. Reversible degenerative chain
transfer is a reaction in which chain-transfer activity is passed on from chain
to chain. Although there are a number of examples of techniques that rely on
this principle (34–37), the most noteworthy undoubtedly is reversible addition–
fragmentation chain transfer (RAFT) (38,39). RAFT was developed by Rizzardo
and co-workers in the late 1990s. Its most popular form makes use of thiocar-
bonylthio compounds that, depending on the activating substituent, can have
very high chain-transfer values. Also, dependent on the activating substituent,
the whole gamut of vinyl monomers can be polymerized in a controlled fashion,
leading to low PDI materials. In contrast to NMP and ATRP, RAFT requires the
continuous production of radicals to keep the polymerization going. Despite this
continuous production of new chains, and the concomitant termination of growing
chains, polymers with PDIs below 1.1 are readily achieved. Also, in this case, a
wide range of polymer architectures has been synthesized.
Simultaneous with the first introduction of RAFT a similar development was
introduced where specifically xanthates were employed as the reversible chain-
transfer agent (40). The technique, called MADIX (macromolecular design via
interchange of xanthates), can be seen as a special case of RAFT polymerization.
Xanthates can be regarded as RAFT agents where the activating substituent is an
alkoxy fragment. For most vinyl monomers, xanthates exhibit a low chain-transfer
constant compared to some of the dithioesters. This leads to broader molar mass
distributions, with PDIs around 2. A notable exception is the combination of vinyl
esters with xanthates. Where it has proven very difficult to perform living poly-
merization of vinyl esters (eg vinyl acetate), the use of xanthates as RAFT agents
Vol. 6
LIVING RADICAL POLYMERIZATION
699
leads to poly(vinyl acetate) of predetermined molar mass, and low polydispersity
(40).
The latest development in living radical polymerization at this stage is the
use of diphenylethylene (DPE) (41). In the continuous search for systems that lead
to living polymerization in combination with a minimal disturbance of the conven-
tional free-radical process, DPE could be an interesting alternative (42). There are
specific difficulties associated with LRP in emulsion (see R
ADICAL POLYMERIZATION
)
(43). DPE, although not fully investigated yet, may prove to be a valuable addition
to the existing LRP techniques.
Nitroxide-Mediated Polymerization (NMP)
Nitroxides have been investigated extensively to control the homo- and copolymer-
ization of styrene. NMP proceeds via the reversible trapping of carbon-centered
radicals as indicated in equation 1. The two most commonly used ways to start
the polymerization are
(1) Generation of free radicals from a conventional free-radical initiator (eg, azo
compound or peroxide), addition of one or a few monomer units, followed by
trapping of the carbon-centered radical by the nitroxide compound (1). The
trapping reaction produces an alkoxy amine, which contains a thermally
labile C O bond (44).
(2) Thermal decomposition of an alkoxy amine (2), which was synthesized in a
separate reaction (45).
N
O
R
R
+
N
O
k
p
1
2
(1)
Since the propagating species in LRP is a free radical, it may undergo the
known reactions, including chain transfer and bimolecular termination. If bi-
molecular termination occurs, the equilibrium will gradually shift in the direction
of the alkoxyamine. This is due to the gradual buildup of nitroxide. This pro-
cess slows down the polymerization, and may lead to unacceptably long reaction
times. Several approaches have been developed to overcome this phenomenon.
In the early days of NMP, TEMPO was by far the most frequently used nitrox-
ide. Because of the small equilibrium constant, the polymerization needed to be
carried out at temperatures above 100
◦
C. Under these conditions, styrene shows
significant thermal (self)initiation. The presence of this additional radical source
counterbalances the buildup of excess free nitroxide. Fukuda and co-workers mea-
sured the rate of TEMPO-mediated styrene polymerization, and found it to be
identical to the conventional thermal polymerization of styrene (46). Workers at
Xerox developed a method to increase the rate of polymerization by the addition
700
LIVING RADICAL POLYMERIZATION
Vol. 6
of camphor sulfonic acid (CSA) (11). Several explanations have been put forward
for the mechanism of rate enhancement by CSA. Most likely the CSA degrades
the excess free nitroxide, thus shifting the equilibrium more to the active species.
An even more effective way to accelerate NMP consists of the addition of a slowly
decomposing free-radical initiator (47,48).
Great improvements in NMP were brought about by the development of new
nitroxides. Two independent groups worked on this topic. Tordo and co-workers de-
veloped
β-phosponylated nitroxides, eg N-tert-butyl-N-(1-diethylphosphono-2,2-
dimethylpropyl)nitroxide, which is commonly known as SG1 (13). Simultane-
ously, Hawker and co-workers synthesized a library of nitroxides and investi-
gated their effectiveness in NMP (12). It turned out that the presence of a
β-
hydrogen greatly improves the performance of the nitroxide. N-tert-butyl-N-(1-
diethylphosphono-2,2-dimethylpropyl)nitroxide and
α-hydrido derivatives based
on a 2,2,5-trimethyl-4-phenyl-3-azahexane-3-oxy skeleton, eg N-tert-butyl-N-(2-
methyl-1-phenyl-propyl)nitroxide, both allow the NMP of acrylates. Further-
more, polymerization temperatures below 100
◦
C are feasible, which greatly fa-
cilitates the use of NMP in emulsion polymerization. This will be discussed
later.
In an NMP initiated by preformed alkoxyamines, the number-average de-
gree of polymerization will be ¯
P
n
= [M]/
[R T]
0
+2[I]
0
, where [R T]
0
is the
alkoxyamine at the start of the reaction,
[I]
0
is the amount of (slowly decom-
posing) initiator converted (dissociated) and
[M] is the amount of monomer con-
verted. The factor 2 associated with the amount of converted initiator accounts for
the fact that one initiator molecule produces two radicals. This equation assumes
the absence of chain-transfer reactions. Note that the occurrence of termination
by disproportionation will not influence the number-average degree of polymer-
ization, whereas termination by combination will lead to a (usually small) devi-
ation from the predicted value. The PDI of the resulting polymer will obviously
be influenced by the occurrence of any chain-breaking reaction (chain transfer or
termination). Details on the prediction of the PDI will be given below.
In the group of Fischer, a considerable amount of work was dedicated to the
kinetic description of NMP. They introduced the concept of the so-called persistent
radical effect (PRE) (49,50). The PRE relies on the limited buildup of the concen-
tration of deactivator, which in the case of NMP is the nitroxide, the persistent
radical. On the basis of the PRE, it can be predicted which conditions will lead to
narrow MMDs, and to a small fraction of irreversibly terminated “dead” chains.
Although many authors present linear relationships between ln ([M]
0
/[M])
and time as one of the proofs that a polymerization is living, Fischer and co-
workers show that ln ([M]
0
/[M]) in many cases should scale with t
2
/3
, and that the
persistent radical concentration increases with t
1
/3
.
In Figure 1 an experimental example from the work of Fischer’s group is
shown that obeys the kinetic relationship they introduced (51). If the kinetic model
is accepted, the implication is a proper prediction of the PDI and the fraction of
dead (irreversibly terminated) chains, etc. In other words, on the basis of kinetic
parameters and relevant concentrations, a system can be optimized in terms of
degree of livingness (low PDI and small fraction of dead chains).
For an overview of rate constants in NMP, the reader is referred to a review
on the persistent radical effect by Fischer (52).
Vol. 6
LIVING RADICAL POLYMERIZATION
701
82.5
°C
3.0
t , 10
3
s
[TEMPO], 10
−
4
M 2.0
1.0
0.0
0
5
10
15
20
25
30
Fig. 1.
Time dependence of the TEMPO concentration during thermolysis of cumyl–
TEMPO:
1.16
× 10
− 2
M;
5.10 × 10
− 3
M;
2.52
× 10
− 3
M.
Atom Transfer Radical Polymerization (ATRP)
ATRP is a more versatile technique than NMP, ie except for a few exceptions
all vinyl monomers are susceptible to ATRP. Notable exceptions are unprotected
acids (eg (meth)acrylic acid). Some other monomers may be difficult to polymerize
since they exhibit side reactions, which may be affected by the choice of reaction
conditions, nature of the catalyst, etc. An example of such a monomer is 4-vinyl
pyridine (4-VP), which can undergo quaternization by the (alkyl halide) initiator
(53). Nevertheless, successful polymerization of 4-VP has been reported.
In ATRP a transition-metal catalyst is used to reversibly activate halide-
terminated dormant chains. The principal mechanism is outlined in equation 2.
R
+
k
p
R
X + Cu
(I)
X/ligand
+ M
Cu
(II)
X
2
/ligand
(2)
The kinetic description of ATRP has received considerable attention. There
is a similarity between the descriptions of NMP and of ATRP. Both rely on re-
versible activation of dormant chains, albeit the activation in the case of NMP is
a unimolecular reaction whereas in the case of ATRP it is a bimolecular reaction.
If the reaction medium is homogeneous, ie activator and deactivator are soluble
in the reaction mixture, the monomer consumption as a function of time can be
described as follows.
702
LIVING RADICAL POLYMERIZATION
Vol. 6
ln
[M]
0
[M]
t
=
k
p
k
a
[RX]
0
[Cu
I
]
0
k
d
[Cu
II
]
0
t
(3)
ln
[M]
0
[M]
t
=
3
2
k
p
k
a
[RX]
0
[Cu
I
]
0
3k
d
×2k
t
1
/3
t
2
/3
(4)
Where [M]
0
and [M]
t
are the monomer concentrations at time zero and time t re-
spectively. [RX]
0
is the alkyl halide initiator concentration at time zero. [Cu
I
]
0
and
[Cu
II
]
0
are the activator and deactivator concentrations, respectively, at time zero.
k
p
is the propagation rate constant, k
a
is the activation rate constant, k
d
is the de-
activation rate constant, and k
t
is the rate coefficient for bimolecular termination.
Equation 3 applies to the situation where a pseudoequilibrium concentration of
deactivator [Cu(II) in this case] is built up. In the case of many ATRP reactions
where no deliberate addition of Cu(II) has taken place, equation 4 will provide
a better match with experimental data. Fischer and co-workers have derived an
expression for the concentration of deactivator necessary to switch from equation
4 to equation 3. The equation derived by Fischer reads (for ATRP):
[Cu(
II
)]
0
,t
=
3K
eq
[RX]
0
[Cu(
I
)]
0
k
t
k
p
(5)
This theoretically derived concentration was experimentally confirmed in a
study by Zhang and co-workers (54). In Figure 2, the experimental data as well as
the theoretical fits to the data are shown from the study by Zhang. It is clear that
the crossover from a t
2
/3
to a linear dependence on time exactly coincides with
the theoretical prediction. In the figure it is obvious that an inhibition period is
observed, which increases with increasing concentration of Cu(II). The origin of
the inhibition period is not completely clear.
In many cases the catalyst complex consists of a Cu(I) halide with multiden-
tate aliphatic tertiary amines as ligands (eg N,N,N
,N
,N
-pentamethyl diethy-
lene triamine, PMDETA). This ligand is cheap and provides relatively high rates of
polymerization with still a reasonable degree of control. One of the peculiarities of
this ligand is the fact that the Cu(II) complex has very limited solubility. In a study
by Snijder and co-workers it is shown by ESR that a ceiling Cu(II) concentration is
reached early in the polymerization of MMA. When this happens, the equations 1
and 2 no longer describe the monomer conversion versus time properly. On the ba-
sis of the experimental observation of constant deactivator concentration, Snijder
derived a new expression that adequately describes monomer conversion versus
time (55).
ln
[M]
0
[M]
=
k
p
[Cu
II
X
2
]
c
2k
tD
k
eq
[3
ξt+[PX]
− 3
0
]
1
/3
−
k
p
[Cu
II
X
2
]
c
2k
tD
k
eq
[PX]
0
with
ξ = 2k
tD
k
eq
/[Cu
II
X
2
]
c
2
.
(6)
Vol. 6
LIVING RADICAL POLYMERIZATION
703
0
100
200
300
400
500
2.0
1.5
1.0
0.5
0.0
(a)
R
2
= 0.9912
R
2
= 0.9931
R
2
= 0.9936
t, min
In([M]
0
/[M])
(b)
R
2
= 0.9981
R
2
= 0.9927
R
2
= 0.9987
1.5
1.0
0.5
0.0
2.0
In([M]
0
/[M])
0
10
20
30
40
50
60
70
t
2/3
, min
2/3
Fig. 2.
Plots of ln([M]
0
/[M]) versus reaction time (t and t
2
/3
) for the ATRP of MMA in
toluene at 90
◦
C using different amounts of initially added CuBr
2
. [MMA]
0
= 2.8879 M,
[HEBIB]
0
= 0.0959 M, [CuBr]
0
= [NHPMI]
0
/3
= 0.0969 M. [CuBr
2
]
0
/[CuBr]
0
:
0;
0.05;
0.1; , 0.2;
0.3.
As indicated above, the majority of work on ATRP has been conducted using
copper as the transition metal. An enormous variety of nitrogen-based ligands
have been used in conjunction with Cu(I) halide. The most popular ligands are
(1) Multidentate aliphatic tertiary amines (PMDETA, Me
6
-TREN).
(2) Substituted bipyridines.
(3) Alkyl pyridylmethanimine.
704
LIVING RADICAL POLYMERIZATION
Vol. 6
The initial work by Matyjaszewski and co-workers largely concentrated on
the use of substituted bipyridines (14). Although this ligand provides good liv-
ing character to the polymerization, the rates of polymerization are quite low. In
more recent work by Matyjaszewski and co-workers as well as in the research
of other groups, more emphasis is placed on the use of multidentate aliphatic
tertiary amines (24). The higher rates of polymerization, and to some extent the
heterogeneous character of the polymerization, have certain advantages, which
will be detailed under Scope and Limitations. The alkyl pyridylmethanimines
were introduced by Haddleton and co-workers (23) and are mainly used for the
polymerization of methacrylates.
Apart from copper-based complexes several other metals have been used as
well. Fe (56–59), Ni (60), Ru (15), etc have been used to some extent. Especially
noteworthy here is the work by Sawamoto and co-workers. As indicated in the In-
troduction, Sawamoto and Matyjaszewski simultaneously pioneered ATRP. Maty-
jaszewski started off with the use of copper, whereas Sawamoto spent most of his
efforts on ruthenium-based catalysts. Recent work from Sawamoto and co-workers
shows that the Ru-based complexes can compete with the Cu-based systems on
many fronts. Although not yet perfect it seems that a specific Fe-based catalyst is
the first to polymerize vinyl acetate via an ATRP mechanism (61).
The initiation in ATRP may occur in one of two different ways. The common
way to initiate is via the reaction of an activated (alkyl) halide with the transition-
metal complex in its lower oxidation state. Typical examples would be the use of
ethyl 2-bromoisobutyrate and a Cu(I) complex for the initiation of a methacrylate
polymerization (62), or 1-phenylethyl chloride for the initiation of a styrene poly-
merization (14). In addition, there are initiators like 2,2,2-trichloro-ethanol (63)
that appear to be very efficient, and that result in hydroxy-functional polymer
chains. Percec and co-workers reported on the use of sulfonyl chlorides as univer-
sal initiators in ATRP (64). Also the use of di-, tri-, or multifunctional initiators is
possible, which will result in polymers growing in two, three, or more directions.
All these variations open up possibilities to well-defined architectures.
The alternative way to initiate ATRP is via a conventional free-radical ini-
tiator, which is used in conjunction with a transition-metal complex in its higher
oxidation state. Typically one would use AIBN in conjunction with a Cu(II) com-
plex. Upon formation of the primary radicals and/or their adducts with a monomer
unit, the Cu(II) complex very efficiently transfers a halogen to this newly formed
chain. In doing so the copper complex is reduced, and the active chain is deacti-
vated. It is easily envisaged that the system will arrive at the same equilibrium
as depicted in equation 2, but now approaching it from the right-hand side. This
alternative way of initiation was termed “reverse ATRP” (65,66).
Reversible Addition–Fragmentation Chain Transfer (RAFT)
In terms of polymerizable monomers, RAFT is at present the most versatile
technique for conducting LRP. In contrast to the previously described NMP and
ATRP, this system relies on chain transfer for the exchange between active and
dormant chains. The chain end of a dormant chain carries a thiocarbonylthio
moiety, which is chain-transfer–active. Upon chain transfer, the thiocarbonylthio
Vol. 6
LIVING RADICAL POLYMERIZATION
705
moiety is transferred to the previously active chain, which now becomes dormant,
and the previously dormant chain carries the radical activity and is able to prop-
agate. In equation 7, the basic mechanism of this exchange reaction via chain
transfer is depicted. As can be seen in equation 7, in between the two states of
dormant and active chains there is an intermediate radical, which will undergo the
subsequent fragmentation reaction (67). In the present literature there is some
controversy about the possible fates of this intermediate radical. Starting point of
the controversy is an observed retardation in some examples of RAFT polymer-
ization compared to the equivalent free-radical polymerization in the absence of
the RAFT agent. Davis and co-workers explain the retardation by a slow rate of
fragmentation, which leads to the buildup of a relatively large concentration of (in-
termediate) radicals (68,69). On the other hand, Monteiro and co-workers explain
the retardation by irreversible termination of the intermediate radical (70). The
remarkable point about this controversy is that the difference in the rate constants
of fragmentation between Davis’ and Monteiro’s approaches is some 7 orders of
magnitude. Fukuda and co-workers carried out independent rate measurements,
which seem to support Monteiro’s view (71). However, at present nobody has ever
been able to detect the presence of three- or four-armed stars, which should result
from the intermediate radical termination.
P
n
S
C
Z
S
+ P
m
P
n
S
C
Z
S
ⴢ
P
m
P
m
S
C
Z
S
+ P
n
(7)
A whole variety of thiocarbonylthio compounds have been synthesized and
used in RAFT polymerization. The initial work was focused to some extent on
the use of dithioesters. More recently the range of RAFT agents is expanded to
trithiocarbonates, dithiocarbamates, and xanthates.
To some extent the choice of RAFT agent determines the degree of control
obtained. The general structure of a RAFT agent can be depicted as in structure
(3), where the Z group is the activating group, and R is the homolytically leaving
group.
S
C
Z
S
R
To a large extent the Z group determines the rate of addition, and the R group
determines the rate of fragmentation. The choice of Z and R groups is dependent
on the nature of the monomer to be polymerized. In Figure 3 an overview is given of
different Z and R groups and the monomers that can be polymerized in a controlled
fashion using these Z and R groups (72).
Although already in the initial patent from CSIRO the application of RAFT
in emulsion is reported (38), it appeared to be quite troublesome to carry out a
well-controlled reaction in emulsion.
706
LIVING RADICAL POLYMERIZATION
Vol. 6
N
Z: Ph >> SCH
3
~ CH
3
~
N
O
>>
> OPh > OEt ~
N
Ph
CH
3
N
Et
Et
>
MMA
STY. MA. AM. AN
VAc
CH
3
CN ~
CH
3
R:
CH
3
Ph >
CH
3
CH
3
COOEt >>
CH
3
CH
3
CH
2
CH
3
CH
3
CH
3
CH
3
~
H
Ph >
CH
3
CH
3
CH
3
~
CH
3
H
Ph
H
MMA
STY. MA. AM. AN. VAc
Fig. 3.
Requirements for Z and R groups of RAFT agent for the controlled polymerization
of styrene (STY), methyl methacrylate (MMA), methyl acrylate (MA), acrylamide (AM),
acrylonitrile (AN), and vinyl acetate (VAc).
Scope and Limitations
Living radical polymerization (LRP) is developed on the basis of conventional
free-radical polymerization (FRP). The main idea behind this development is to
overcome intrinsic shortcomings of FRP without sacrificing any of its strengths.
These strengths are
(1) Wide range of monomers, including functional ones (hydroxyl, epoxy, car-
boxylic acid, etc.).
(2) Nondemanding in terms of reaction conditions.
(3) Widely applied in academia as well as in industry for (co)polymer synthesis.
The intrinsic shortcomings are
(1) Poor control over chain length (distribution).
(2) Poor control over chain topology (block copolymers and more advanced
topologies are virtually impossible to synthesize efficiently).
(3) No, or hardly any, control over tacticity.
To what extent did the efforts in LRP up till now lead to success in overcoming
the shortcomings of FRP? The one shortcoming that has only been solved very
recently is the control over tacticity. Matyjaszewski and co-workers have recently
shown that the addition of lanthanides to an ATRP results in control over tacticity
(73).
Various classes of monomers have been polymerized via LRP. Styrene
and its derivatives can be polymerized using any of the three main tech-
niques. Under the appropriate conditions NMP, ATRP, and RAFT all lead to
polymers with a predetermined molar mass, and a low PDI. Also the syn-
thesis of block copolymers where one of the blocks is poly(STY) has success-
fully been accomplished. Styrene is the monomer that poses the smallest de-
mands on the choice of nitroxide in NMP. TEMPO has been used frequently, but
Vol. 6
LIVING RADICAL POLYMERIZATION
707
one requires relatively high reaction temperatures (125–140
◦
C). When N-tert-
butyl-N-(1-diethylphosphono-2,2-dimethylpropyl)nitroxide or N-tert-butyl-N-(2-
methyl-1-phenyl-propyl)nitroxide is used, the reaction temperature can be low-
ered to 80–90
◦
C. Similarly, in ATRP a wide variety of transition-metal complexes
can be used in combination with a large selection of initiators. Styrene appears to
posses an equilibrium constant for the activation–deactivation equilibrium, which
is quite favorable for conducting these reactions. Finally, for RAFT also, the choice
of RAFT agent is not very critical in the degree of success obtained. Usually, large
chain-transfer rate constants are obtained, which results in polymers with low
PDIs.
Methacrylates form an important class of monomers. They are available with
a large number of different ester side groups. NMP is not capable of polymerizing
methacrylates in a controlled fashion. Most frequently the nitroxide will undergo
a disproportionation reaction, with the propagating radical leading to a dead poly-
mer chain with an unsaturated chain end and a hydroxylamine.
ATRP can in general be carried out on methacrylates. The initial work on
MMA polymerization was reported by the group of Matyjaszewski (62), includ-
ing work on high molar mass polymer (74), detailed kinetics (75), and on sub-
stituted methacrylates (76–78). Haddleton and co-workers focused largely on the
polymerization of methacrylates, and used most frequently alkyl pyridylmetha-
nimine ligands with copper halide as the catalyst. The only occasion where ATRP
on methacrylates leads to difficulties is when the ester side chain contains func-
tional groups that interfere with the copper complex. Many other groups have
investigated the polymerization of methacrylates. A variety of transition-metal
complexes have been utilized for this purpose. The parent compound of methacry-
lates (methacrylic acid) cannot be polymerized as such via ATRP. Armes and co-
workers showed that the sodium salt of methacrylic acid can be polymerized in
aqueous media (79). Other groups used approaches where the acid functionality
is protected via various methodologies, eg via tert-butyl ester, via trimethylsilyl
ester, or via the tetrahydropyranyl ester (80).
RAFT is probably the most versatile technique to polymerize methacrylates.
Regardless of the nature of the ester side group, RAFT can be employed for the LRP
of methacrylates. Also, methacrylic acid can be polymerized without protecting
group chemistry. There are some requirements in terms of activating Z group and
homolytically leaving R group of the RAFT agent that have to be met in order to
have a successful polymerization.
Acrylates can be polymerized using NMP when the proper nitroxide is em-
ployed. N-tert-butyl-N-(1-diethylphosphono-2,2-dimethylpropyl)nitroxide and N-
tert-butyl-N-(2-methyl-1-phenyl-propyl)nitroxide were proven to be capable of
controlling the polymerization of butylacrylate.
ATRP of acrylates is easily achieved. The equilibrium between active and
dormant chains is much more to the dormant side. This results in relatively low
rates of polymerization even though the propagation rate constant of acrylates is
quite large. Methacrylate polymerizations, although having a much smaller prop-
agation rate constant, proceed faster in ATRP than do acrylate polymerizations
under identical conditions. In general, polyacrylates will posses a somewhat larger
PDI than polymethacrylates synthesized via ATRP, especially at low molar mass.
The main reason of this phenomenon is that during an activation–deactivation
708
LIVING RADICAL POLYMERIZATION
Vol. 6
cycle acrylates will undergo multiple monomer additions, whereas methacrylates
will only undergo single monomer addition. In the long chain limit this effect will
diminish. Also, in the case of acrylates, acrylic acid can only be polymerized in the
nonprotonated form (sodium salt or protected as indicated above).
Similar to the polymerization of methacrylates, acrylates can easily be poly-
merized using RAFT. In this case the choice of RAFT agent is less critical, and
functional groups (including acid) are not known or expected to result in any prob-
lems. The only point of caution is that depending on the choice of the RAFT agent
significant retardation or inhibition may occur.
Butadiene and isoprene have to date only been polymerized in a living radical
fashion via NMP (81). The group of Hawker has shown that nitroxides with an
alpha C H bond to the nitrogen can successfully be applied in the polymerization
of dienes. The result is a linear increase of molar mass with conversion, and a PDI
below 1.1.
Vinyl acetate is the monomer that has required significant efforts in order
to achieve LRP. The first successful attempt was using RAFT with xanthates as
a RAFT agent (also known as MADIX). At a later stage it turned out that certain
dithiocarbamates also were able to induce LRP of vinyl acetate. Of all the attempts
to polymerize VAc using ATRP, only specific iron complexes used by Sawamoto and
co-workers were claimed to be successful (61). However, the PDIs they claimed
were relatively large compared to other monomer/catalyst combinations.
Although the synthesis of polymers with narrow molar mass distributions
is interesting from an academic point of view, the potential applications are more
likely to be found in polymers with specific topologies. The most basic topology in
reach of LRP is the block copolymer. Although it may occur as a trivial exercise
to synthesize an A–B block copolymer, there are some precautions that need to
be taken into account. The trivial one is that a maximum degree of livingness
should be obtained for the first block. In general this requires some limitations
to the maximum conversion in the synthesis of the first block. The probability
of bimolecular termination significantly increases if the polymerization would
be carried out to (almost) complete monomer conversion. The other important
precaution concerns the sequence of block copolymer synthesis. Whether A–B or
B–A is the preferred sequence mainly depends on the rate of initiation of the
second monomer by the first block macroinitiator. As an example, when poly(STY-
b-MMA) is the target structure, it is either the initiation of STY polymerization
by the poly(MMA) macroinitiator or the initiation of MMA polymerization by the
poly(STY) macroinitiator. In the case of ATRP as the LRP technique of choice, and
without the use of any advanced techniques (which will be outlined below) initi-
ation versus propagation needs to be studied. For the proper synthesis of a block
copolymer a clean transition is required. Hence, initiation should be fast compared
to propagation. When poly(STY) is used to initiate MMA this requirement is not
met. Initiation is relatively slow, since for poly(STY) the activation–deactivation
equilibrium lies strongly to the dormant side. To some extent this means that each
chain that has undergone the transition to block copolymer is more reactive than
the poly(STY) macroinitiator. The former chains will start to grow with a rela-
tively high rate whereas the remaining macroinitiators still wait to be activated.
The result is a broad molar mass distribution, or in the case of a low molar mass
poly(MMA) block, a large fraction of unreacted poly(STY) macroinitiator.
Vol. 6
LIVING RADICAL POLYMERIZATION
709
Conversely, when poly(MMA) is used as a macroinitiator, initiation is rela-
tively fast since the equilibrium between active and dormant chains lies more to
the active side. The chains that get converted into block copolymer are less reac-
tive in this case. In the limit it can be explained as a situation in which all the
macroinitiators get converted into block copolymer before the chain growth of the
second block [poly(STY)] commences.
As indicated above, this scenario holds for ATRP when no special techniques
are invoked. There is one relatively straightforward technique to circumvent
this forced order of block copolymer synthesis. This technique is called halogen-
exchange (83). It relies on the difference in bond strength between an alkyl chlo-
ride and an alkyl bromide. The carbon–bromine bond is weaker than the carbon–
chlorine bond. The result of this in a so-called mixed halogen system is that
carbon–chlorine bonds are preferentially formed. In halogen exchange experi-
ments, this phenomenon is used to slow down propagation relative to initiation.
In practice this would mean that poly(STY) is synthesized from a bromine func-
tional initiator using copper(I) bromide as the catalyst. In the second step, where
MMA is polymerized, the bromine functional poly(STY) macroinitiator is used in
conjunction with copper(I) chloride. The macroinitiator chain gets activated and
adds an MMA unit. Subsequently, it is deactivated by CuCl
2
(or CuClBr), which
results in the formation of a carbon - chlorine bond, and thus a chlorine func-
tional dormant block copolymer chain. Because of the difference in bond strength,
the poly(STY) Br is now more reactive than the poly(STY-b-MMA) Cl. Hence,
all the macroinitiator gets converted into dormant block copolymer chains, fol-
lowed by propagation of the MMA block. Initially, the work by Matyjaszewski
and co-workers on halogen exchange strongly emphasized the negative effect on
propagation rate. This obviously leads to an increased ratio of initiation rate to
propagation rate. At a later stage it was shown by Klumperman and co-workers
that, apart from the effect on propagation rate, the use of copper chloride also leads
to an increased rate of initiation (larger value of the activation rate constant of a
bromine functional macroinitiator).
In Figure 4 it is shown how halogen exchange in block copolymer synthesis
influences the resulting molar mass distributions (82).
Advanced Architectures
As indicated above, conventional free-radical polymerization is not very well
suited for the synthesis of advanced polymer architectures. Even the synthesis
of di- or triblock copolymers is hardly possible using this technique. The main
reason is the continuous initiation and termination of chains. This results in the
early generation of dead polymer chains that no longer participate in the poly-
merization process. The consecutive addition of monomers as employed in living
anionic polymerization for the synthesis of eg poly(styrene-block-butadiene) will
only lead to the synthesis of a mixture of two homopolymers.
Advanced architectures are possible via LRP. A large fraction of the chains in
an LRP will grow throughout the polymerization process. The subsequent addition
of a second monomer will lead to the formation of a block copolymer. Various
examples of this process can be found in recent literature employing any of the
710
LIVING RADICAL POLYMERIZATION
Vol. 6
Diblock copolymer
PBA
1
0.8
0.6
0.4
0.2
0
1000
10000
100000
RI signal
Molecular weight
(a)
% Conversion
= 0
PMA-Cl
PMA-
b
-PMMA
29
45
58
72
1
0.8
0.6
0.4
0.2
0
1000
10000
100000
1000000
Molecular weight
(b)
RI signal
Fig. 4.
Effect of halogen exchange on the molar mass distribution of an acrylate–
methacrylate diblock copolymer. (a) Poly(BA-b-MMA) with halogen exchange; (b). Poly(MA-
b-MMA) without halogen exchange (82).
known techniques ATRP, NMP, or RAFT (84,85). As indicated above, the sequence
of the synthesis of a block copolymer may have a large influence on the degree of
control.
The synthesis of an A B C triblock copolymer can be realized using an anal-
ogous procedure (86). Also, in this case the sequence of the comonomer additions
Vol. 6
LIVING RADICAL POLYMERIZATION
711
will influence the final result. However, in this case monomer B needs to be in-
corporated as the second block. Smart strategies, which include the proper choice
of comonomer sequence and the possibility to employ halogen exchange in ATRP,
are needed to optimize the resulting block copolymer. If the triblock copolymer is
a symmetrical A–B–A copolymer, the most commonly employed synthetic strat-
egy is the use of a difunctional initiator (NMP or ATRP) or transfer agent (RAFT).
This allows the growth of the central B-block in two directions, and the subsequent
addition of monomer A to synthesize the outer blocks. Here obviously the same
precautions and limitations apply as in the synthesis of A–B diblock copolymers.
One of the more or less unique features of LRP in the synthesis of block
copolymers is the possibility to incorporate a statistical copolymer as one of the
blocks. Because of the very favorable reactivity ratios in radical polymerization
compared to ionic polymerizations this is more easily achieved in LRP than in eg
anionic polymerization. If the statistical copolymer is synthesized at its azeotropic
conditions, there is little difference with the synthesis of a homopolymer block. If
polymerization would take place under nonazeotropic conditions, the usual pre-
cautions need to be taken to avoid composition drift.
On the other hand, composition drift can be employed in LRP to synthesize
gradient (or tapered) copolymers (87). The prediction of the composition drift can
be carried out in exact analogy with composition drift in conventional free-radical
copolymerization. The only, but important, difference is that in the case of LRP,
the composition drift is superimposed on the growth of each individual chain. This
then automatically leads to the formation of a gradient copolymer. In extraordi-
nary cases the composition drift is such that a spontaneous block copolymerization
will take place. One example that has been published in this regard is the synthe-
sis of a poly(styrene-block-(styrene-alt-maleic anydride)). When a polymerization
is started at a STY to MAnh molar ratio of 75:25, a copolymer with close to al-
ternating monomer sequence is formed. This leads to a relatively fast depletion
of MAnh, with the excess STY forming a homopolymer block. For the synthesis of
alternating copolymers it turns out that the commonly used strategy in FRP of
adding a Lewis acid to the copolymerization of styrene and acrylates can also be
employed in ATRP (88)
Another class of architectures in reach of LRP are the so-called hyper-
branched structures (see H
YPERBRANCHED
P
OLYMERS
). The well known limit of
these hyperbranched structures are dendrimers, where the consecutive growth
of the polymer leads to (nearly) perfect globular macromolecules. The synthesis
of dendrimers usually is a tedious multistep organic reaction. In the case of LRP,
where monomers are polymerized that carry an initiator functionality as well, the
synthetic procedure is quite straightforward. Examples have been reported on the
basis of NMP (89) as well as on the basis of ATRP (90,91). It is very important to
tailor the reactivities of initiator and monomer in this application. Two limiting
cases can be recognized in that sense. If initiation is slow compared to propaga-
tion, linear chains will be formed that carry the initiator functionalities as pen-
dent groups. In principle these polymers could be used for the synthesis of densely
grafted polymers as described below. On the other hand, if initiation is very fast
compared to propagation, monomers will be enchained in linear chains as well, but
the polymer backbone now has a completely different architecture. In Figure 5,
the two modes of polymerization are illustrated using para-vinylbenzylchloride.
712
LIVING RADICAL POLYMERIZATION
Vol. 6
CH
2
Cl
CH
2
Cl
CH
2
Cl
CH
2
Cl
CH
2
Cl
CH
2
Cl
(a)
CH
2
Cl
Cl
Cl
(b)
Fig. 5.
Two modes of polymerization of para-vinylbenzylchloride, resulting from (a) fast
propagation compared to initiation, and from (b) fast initiation compared to propagation.
R
divinyl
compound
X
X
R
R
R
polymer
linking
star—star
coupling
Fig. 6.
Synthesis of star polymers via the “arm first” strategy (94).
It needs to be stressed that Figure 5 shows the hypothetical two extreme cases
outlined above.
The first examples of star polymers synthesized via LRP were reported by
Matyjaszewski and co-workers (92,93). At a later stage they reported on the syn-
thesis of star polymers via the so-called arm first procedure (94). In this approach
polymer chains are synthesized with a reactive chain end that allows coupling to
the core of the star (Fig. 6).
The statistical nature of the growth process of the hyperbranched struc-
tures will inevitably lead to a relatively large variation in degree of branching.
The closest one could get to a dendritic structure is by carefully matching the
initiation and propagation rates. Here it needs to be taken into account that in
the example shown in Figure 5, the initial primary halide will be turned into
a secondary halide after the first monomer addition. This will greatly affect its
activation–deactivation equilibrium. Alternatives to true hyperbranched struc-
tures in some applications are star-branched structures with a fixed and known
number of arms. In LRP these structures are readily accessible via multifunctional
Vol. 6
LIVING RADICAL POLYMERIZATION
713
initiators or transfer agents. Examples have been published using NMP, but more
notably using ATRP. In the work of Matyjaszewski it has been shown that any
hydroxy functionality can be turned into an ATRP initiator via esterification using
2-bromo-2-methyl-propionyl bromide, as indicated in equation 8 (95).
OH + Br
CH
3
CH
3
Br
O
Br
CH
3
CH
3
O
O
(8)
This approach was used by Haddleton and co-workers to synthesize multi-
functional initiators and subsequently multiarm stars by using various sugars
(96), and cyclodextrin (21-arm star) (97). The general advantages of LRP tech-
niques also apply to the synthesis of the multiarm stars. This includes the possi-
bility to synthesize block copolymer arms, gradient polymer arms, etc. One of the
potential drawbacks is the enhanced tendency toward bimolecular termination,
which in the case of termination by coupling leads to star–star coupling. It will be
evident that the step from star–star coupling to network formation or cross-linking
(gelation) is only a relatively minor one.
The final architecture that will be discussed here is that of densely grafted
polymer chains. If the process described for the formation of hyperbranched
structures is turned into a two-step process, it is possible to synthesize chains
that contain a branch on each repeat unit of the polymer backbone. Because
of the great steric congestion, these structures will behave like a rigid rod. The
first example of a densely grafted polymer chain completely via LRP (ATRP) is
based on the polymerization of trimethylsilyl-protected 2-hydroxyethyl methacry-
late (HEMA-TMS) (77). In a post-polymerization reaction the trimethylsilyl
groups are removed and the resulting hydroxy functionalities are turned into
ATRP initiators by reaction with 2-bromoisobutyryl bromide to yield poly-(2-
(2-bromoisobutyryloxy)ethyl methacrylate) (pBIEM). Subsequently, the densely
grafted structure was obtained by polymerizing styrene or butyl acrylate as side
chains from the initial backbone (pBIEM). In Figure 7 typical examples of these
rigid rod structures are shown as observed by Atomic Force Microscopy (qv)
(AFM).
Combined Mechanisms
LRP is a powerful tool for the synthesis of complex polymer architectures as was
shown above. However, in some cases it is desirable to combine structures that
are hardly or not at all accessible via radical polymerization techniques. In such
cases it may be beneficial to combine LRP with another polymerization mecha-
nism. Many examples have been reported so far. A few examples will be listed
here. Polystyrene-b-polyisobutylene-b-polystyrene was synthesized via a combi-
nation of living cationic polymerization and ATRP (98). Polyolefin Graft Copoly-
mers (qv) were synthesized by first polymerizing alkoxyamine-substituted olefins
via metallocene catalysis, and subsequent polymerization of vinyl monomers via
714
LIVING RADICAL POLYMERIZATION
Vol. 6
200
400
600
m4
Fig. 7.
AFM image of a densely grafted copolymer. Polystyrene grafted from pBIEM
(M
n
= 55,500; M
w
/M
n
= 1.3).
NMP (99). Living ring-opening polymerization of
ε-caprolactone and lactides was
combined with various LRP techniques (100). An approach to synthesize vinyl
polymers with poly(vinyl ether) grafts is to make use of macromonomers. Cationic
polymerization is used to synthesize the macromonomers with a methacryloyl
pendent group. The macromonomers are then copolymerized via ATRP (101).
Poly(
ε-caprolactone)-graft-poly(methyl methacrylate) copolymers are prepared by
the (co)polymerization of a cyclic ester containing an activated bromide functional
group [
γ -(2-bromo-2-methylpropionyloxy)-ε-caprolactone]. The activated bromide
functionality is subsequently used to introduce the PMMA side chains via ATRP
(102). Anionic Polymerization (qv) and ATRP are combined to synthesize eg A–B–C
triblock copolymers (103). The same group demonstrated the versatility of end-
group modification of various polymers into ATRP initiators (104,105). It needs to
be stressed that this approach has been adopted by many research groups, and that
probably the first attempts were published by Matyjaszewski and co-workers (18).
Polyisoprene-b-polystyrene with a single fluorescent label at the junction between
the two blocks is reported by Winnik and co-workers (106). They also use a com-
bination of anionic polymerization and ATRP. Finally, LRP can be combined with
enzymatic polymerization as shown by Meyer and co-workers (107). They use a bi-
functional initiator (benzyl-3-(2-bromo-2-methylpropionyloxy)-2-hydroxymethyl-
2-methylpropionate) with a primary hydroxy functionality to initiate the enzy-
matic ring-opening polymerization of
ε-caprolactone, and an activated bromide
functionality to initiate ATRP.
Vol. 6
LIVING RADICAL POLYMERIZATION
715
LRP in Emulsion
One of the great advantages of conventional FRP is the fact that the polymer-
ization can be conducted in aqueous environment, viz suspension polymerization,
emulsion polymerization (see H
ETEROPHASE
P
OLYMERIZATION
). Soon after LRP was
first published, investigations were started to allow LRP to be carried out in aque-
ous media as well. This turned out to be far from trivial. Recently, the efforts to
conduct LRP in emulsion were reviewed (108).
Bon and co-workers were the first to publish a successful attempt toward
TEMPO-mediated LRP in emulsion (109). This first work was far from optimal,
since it only worked in seeded emulsion polymerization, and temperatures above
100
◦
C were required to obtain a dynamic equilibrium between dormant and ac-
tive species. NMP improved quite significantly with respect to its applicability
in emulsion polymerization when the above-mentioned developments led to poly-
merization temperatures below 100
◦
C. Notable in this respect was the work by
Charleux and co-workers, who employed N-tert-butyl-N-(1-diethylphosphono-2,2-
dimethylpropyl)nitroxide as the mediating nitroxide for the LRP of styrene and
butyl acrylate in emulsion (110,111).
ATRP in emulsion (112–116) and miniemulsion (117) was mainly investi-
gated by Matyjaszewski and co-workers. The difficulties to remove the copper
catalyst from the latex, and the low stability of the obtained latex were among
the structural difficulties to develop an attractive alternative to the NMP-based
systems.
RAFT polymerization in emulsion (118–121) and miniemulsion (122–125)
received considerable attention. The initial results were not very promising. In
many instances the stability of the emulsion during polymerization was very poor.
Phase separation was often observed where the typical red color of the RAFT
agent would disappear from the emulsion and form a sticky red layer on top of the
emulsion. One of the alternatives to overcome this problem was the application
of miniemulsion polymerization. In this technique, micron-size monomer droplets
are absent. All the monomer is dispersed in submicron droplets, which are trans-
formed into latex particles during the polymerization process. Stable latexes were
obtained in a number of cases, where the polymers possess a reasonably narrow
molar mass distribution. The disadvantage of miniemulsion polymerization is that
relatively high surfactant concentrations and the use of a so-called cosurfactant
or hydrophobe is required. It is therefore that further research was conducted
to develop a strategy that allows LRP in conventional emulsion. At present the
most successful attempt seems to be a two-stage process (126) in which a water-
soluble monomer, eg acrylic acid, is polymerized in the presence of a RAFT agent
in aqueous media. This polymerization is only conducted up to low molar mass.
Subsequently, a monomer with low water solubility is dispersed in the reaction
mixture, and further polymerization is carried out to yield a stable latex. For-
mally the polymer is a diblock copolymer, albeit with a short hydrophilic block.
The polymer has a relatively low PDI.
The use of RAFT agents with a low activity (120,121) ie a low chain-transfer
constant is another option that provides useful results. Xanthates represent a class
of RAFT agents that are useful in this respect. The majority of polymer chains
in such a system carry the xanthate as an end group. However, because of the
716
LIVING RADICAL POLYMERIZATION
Vol. 6
low chain-transfer constant the PDI of the polymer is around 2. In principle, the
polymers can be used for chain extension, and for the synthesis of block copolymers
(qv).
The final system that is worth mentioning in this chapter on LRP in emul-
sion is the use of 1,1-diphenylethylene (DPE) (127). DPE adds to a growing radical
and forms a radical with a reactivity too low for propagation. The exact mecha-
nism is not elucidated, but the incorporation of DPE leads to a chain that allows
chain extension or block copolymer formation. More or less similar to what was
described about the xanthates in RAFT polymerization, here also the molar mass
distributions are relatively broad. The greatest advantage of DPE-mediated poly-
merization is the fact that it results in minimal disturbance of the polymerization
process. The product latex does not contain any unusual extractable material
(like the catalyst in ATRP), or polymer-bound colored moieties (like the thiocar-
bonylthio compound in RAFT). The obvious drawback is the limited control over
chain architecture, and the limited understanding of the mechanistic details.
Future Directions
Presently, significant work has been done to develop and optimize various LRP
techniques. Commercial applications are still scarce. The general expectation is
that the products of LRP will mainly be employed in high-value-added products.
These may include specialty coatings, adhesives, personal care products, biomedi-
cal applications, and optoelectronical applications. Examples of biomedical appli-
cations can be found in the work of eg, M ¨
uller (128), Haddleton (129), and Wooley
(130,131). For materials in optoelectronical applications it is of great importance
that many of the techniques from FRP can be adopted to LRP. An example is
surface grafting from or onto various substrates, like silica (132) and gold (133).
Much of the work of Hawker and co-workers is dedicated to the application of LRP
in electronical applications (134).
BIBLIOGRAPHY
“Living Polymer Systems, Radical Polymerization” in EPSE 2nd ed., Suppl. Vol., pp. 429–
436, by H. J. Harwood, University of Akron.
1. G. Odian, Principles of Polymerization, 3rd ed., John Wiley & Sons, Inc., New York,
1991.
2. M. Swarc, Adv. Polym. Sci. 2, 275 (1960).
3. E. Borsig, M. Lazar, M. Capla, and S. Florian, Angew. Makromol. Chem. 9, 89 (1969).
4. T. Otsu, M. Yoshida, and T. Tazaki, Makromol. Chem., Rapid Commun. 3, 133–140
(1982).
5. T. Otsu and T. Tazaki, Polym. Bull. 16, 277–84 (1986).
6. T. Otsu, K. Yamashita, and K. Tsuda, Macromolecules 19, 287–290 (1986).
7. T. Otsu and M. Yoshida, Polym. Bull. 7, 197–203 (1982).
8. D. R. Suwier, M. N. Teerenstra, B. Vanhaecht, and C. E. J. Koning, Polym. Sci., Part
A: Polym. Chem. 38, 3558–3568 (2000).
9. Eur. Pat. EP 135280 A2 (1985), D. H. Solomon, E. Rizzardo, and P. Cacioli (Common-
wealth Scientific and Industrial Research Organization).
Vol. 6
LIVING RADICAL POLYMERIZATION
717
10. M. K. Georges, R. P. N. Veregin, G. K. Hamer, and P. M. Kazmaier, Macromol. Symp.
88, 89–103 (1994).
11. M. K. Georges, R. A. Kee, R. P. N. Veregin, G. K. Hamer, and P. M. Kazmaier, J. Phys.
Org. Chem. 8, 301–305 (1995).
12. D. Benoit, V. Chaplinski, R. Braslau, and C. J. Hawker, J. Am. Chem. Soc. 121, 3904–
3920 (1999).
13. D. Benoit, S. Grimaldi, S. Robin, J.-P. Finet, P. Tordo, and Y. Gnanou, J. Am. Chem.
Soc. 122, 5929–5939 (2000).
14. J.-S. Wang and K. Matyjaszewski, J. Am. Chem. Soc. 117, 5614–5615 (1995).
15. M. Kato, M. Kamigaito, M. Sawamoto, and T. Higashimura, Macromolecules 28, 1721–
1723 (1995).
16. D. P. Curran, in B. M. Trost and I. Fleming, eds., Comprehensive Organic Synthesis,
Vol. 4, Pergamon, Oxford, UK, 1991, p. 715.
17. V. Coessens, T. Pintauer, and K. Matyjaszewski, Prog. Polym. Sci. 26, 337 (2001).
18. J.-S. Wang, D. Geszta, and K. Matyjaszewski, Polym. Mater. Sci. Eng. 73, 416 (1995).
19. K. Matyjaszewski, S. Coca, and Y. Nakagawa, J. Xia Polym. Mater. Sci. Eng. 76, 147
(1997)
20. C. Waterson and D. M. Haddleton, Polym. Prepr. 40, 1045–1046 (1999).
21. D. M. Haddleton, D. Kukulj, E. J. Kelly, and C. Waterson, Polym. Mater. Sci. Eng. 80,
145–146 (1999).
22. S. A. F. Bon, A. G. Steward, and D. M. Haddleton, J. Polym. Sci., Part A: Polym. Chem.
38, 2678–2686 (2000).
23. D. M. Haddleton, M. C. Crossman, B. H. Dana, D. J. Duncalf, A. M. Heming, D. Kukulj,
and A. J. Shooter, Macromolecules 32, 2110–2119 (1999).
24. J. Xia and K. Matyjaszewski, Macromolecules 30, 7697–7700 (1997).
25. J. Xia, S. G. Gaynor, and K. Matyjaszewski, Macromolecules 31, 5958–5959 (1998).
26. H. J. Paik, S. G. Gaynor, and K. Matyjaszewski, Macromol. Rapid Commun. 19, 47–52
(1998).
27. D. Mecerreyes, G. Moineau, P. Dubois, R. Jerome, J. L. Hedrick, C. J. Hawker, E. E.
Malmstrom, and M. Trollsas, Angew. Chem., Int. Ed. 37, 1274–1276 (1998).
28. T. Fonagy, B. Ivan, and M. Szesztay, Polym. Mater. Sci. Eng. 79, 3–4 (1998).
29. K. L. Beers, S. G. Gaynor, K. Matyjaszewski, S. S. Sheiko, and M. Moeller, Macro-
molecules 31, 9413–9415 (1998).
30. S. C. Hong and K. Matyjaszewski, Macromolecules 35, 7592–7605 (2002).
31. S. Angot, N. Ayres, S. A. F. Bon, and D. M. Haddleton, Macromolecules 34, 768–774
(2001).
32. A. J. Carmichael, D. M. Haddleton, S. A. F. Bon, and K. R. Seddon, Chem. Commun.
1237–1238 (2000).
33. D. M. Haddleton, S. G. Jackson, and S. A. F. Bon, J. Am. Chem. Soc. 122, 1542–1543
(2000).
34. K. Matyjaszewski, S. G. Gaynor, and J.-S. Wang, Macromolecules 28, 2093 (1995).
35. D. M. Haddleton, C. Topping, D. Kukulj, and D. Irvine, Polymer 39, 3119–3128
(1998).
36. C. L. Moad, G. Moad, E. Rizzardo, and S. H. Thang, Macromolecules 29, 7717–7726
(1996).
37. S. Jiang, H. G. Viehe, N. Oger, and D. Charmot, Macromol. Chem. Phys. 196, 2349–
2360 (1995).
38. PCT Int. Appl. WO 9801478 A1 (1998), T. P. Le, G. Moad, E. Rizzardo, and S. H. Thang
(Commonwealth Scientific and Industrial Research Organization).
39. J. Chiefari, Y. K. Chong, F. Ercole, J. Krstina, J. Jeffery, T. P. T. Le, R. T. A. Mayadunne,
G. F. Meijs, C. L. Moad, G. Moad, E. Rizzardo, and S. H. Thang, Macromolecules 31,
5559–5562 (1998).
718
LIVING RADICAL POLYMERIZATION
Vol. 6
40. PCT Int. Appl. WO 9858974 A1 (1998), P. Corpart, D. Charmot, T. Biadatti, S. Zard,
and D. Michelet (Rhodia).
41. P. C. Wieland, B. Raether, and O. Nuyken, Macromol. Rapid Commun. 22, 700–703
(2001).
42. S. Viala, K. Tauer, M. Antonietti, R.-P. Kruger, and W. Bremser, Polymer 43, 7231–
7241 (2002).
43. J. Qiu, B. Charleux, and K. Matyjaszewski, Prog. Polym. Sci. 26, 2083 (2001).
44. C. J. Hawker, G. G. Barclay, A. Orellana, J. Dao, and W. Devonport, Macromolecules
29, 5245–5254 (1996).
45. G. Moad and E. Rizzardo, Macromolecules 28, 8722–8788 (1995).
46. T. Fukuda, T. Terauchi, A. Goto, K. Ohno, Y. Tsujii, T. Miyamoto, S. Kobatake, and B.
Yamada, Macromolecules 29, 6393–6398 (1996).
47. A. Goto and T. Fukuda, Macromolecules 30, 4272 (1997).
48. D. Greszta and K. J. Matyjaszewski, Polym. Sci., Part A: Polym. Chem. 35, 1857 (1997).
49. H. Fischer, J. Polym. Sci., Part A: Polym. Chem. 37, 1885–1901 (1999).
50. H. Fischer, Macromolecules 30, 5666–5672 (1997).
51. T. Kothe, S. Marque, R. Martschke, M. Popov, and H. Fischer, J. Chem. Soc., Perkin
Trans. 2, 1553–1559 (1998).
52. H. Fischer, Chem. Rev. 101, 3581–3610 (2001).
53. H. B. Sonmez and N. Bicak, React. Funct. Polym. 51, 55–60 (2002).
54. H. Zhang, B. Klumperman, W. Ming, H. Fischer, and R. van der Linde, Macromolecules
34, 6169–6172 (2001).
55. A. Snijder, B. Klumperman, and R. van der Linde, Macromolecules 35, 4785–4790
(2002).
56. G. Moineau, Ph. Dubois, R. Jerome, T. Senninger, and Ph. Teyssie, Macromolecules
31, 545–547 (1998).
57. S. Zhu, D. Yan, and G. Zhang, Polym. Bull. 45, 457–464 (2001).
58. K. Matyjaszewski, M. Wei, J. Xia, and N. E. McDermott, Macromolecules 30, 8161
(1997).
59. M. Teodorescu, S. G. Gaynor, and K. Matyjaszewski, Macromolecules 33, 2335 (2000).
60. G. Moineau, M. Minet, P. Teyssie, and R. Jerome, Macromol. Chem. Phys. 201, 1108–
1114 (2000).
61. M. Wakioka, K.-Y. Baek, T. Ando, M. Kamigaito, and M. Sawamoto, Macromolecules
35, 330–333 (2002).
62. M. Destarac, K. Matyjaszewski, and B. Boutevin, Macromol. Chem. Phys. 201, 265–
272 (2000).
63. V. Percec, H.-J. Kim, and B. Barboiu, Macromolecules 30, 8526–8528 (1997).
64. J. S. Wang and K. Matyjaszewski, Macromolecules 28, 7572 (1995).
65. J. Xia and K. Matyjaszewski, Macromolecules 30, 7692 (1997).
66. A. Goto, K. Sato, Y. Tsujii, T. Fukuda, G. Moad, E. Rizzardo, and S. H. Thang, Macro-
molecules 34, 402–408 (2001).
67. S. Perrier, C. Barner-Kowollik, J. F. Quinn, P. Vana, and T. P. Davis, Macromolecules
35, 8300–8306 (2002).
68. P. Vana, T. P. Davis, and C. Barner-Kowollik, Macromol. Theory Simul. 11, 823–835
(2002).
69. M. J. Monteiro and H. de Brouwer, Macromolecules 34, 349–352 (2001).
70. A. Goto, K. Sato, Y. Tsujii, T. Fukuda, G. Moad, E. Rizzardo, and S. H. Thang, Macro-
molecules 34, 402–408 (2001).
71. S. H. Thang, personal communication, 2001.
72. J. F. Lutz, D. Neugebauer, and K. Matyjaszewski, J. Am. Chem. Soc. (submitted).
73. J.-S. Wang and K. Matyjaszewski, Macromolecules 28, 7901 (1995).
74. T. Grimaud and K. Matyjaszewski, Macromolecules 30, 2216 (1997).
Vol. 6
LIVING RADICAL POLYMERIZATION
719
75. J.-L. Wang, T. Grimaud and K. Matyjaszewski, Macromolecules 30, 6507 (1997).
76. X. Zhang, J. Xia, and K. Matyjaszewski, Macromolecules 31, 5167 (1998).
77. K. L. Beers, S. G. Gaynor, K. Matyjaszewski, S. S. Sheiko, and M. Moeller, Macro-
molecules 31, 9413 (1998).
78. K. L. Beers, S. Boo, S. G. Gaynor, and K. Matyjaszewski, Macromolecules 32, 5772
(1999).
79. X.-S. Wang, R. A. Jackson, and S. P. Armes, Macromolecules 33, 255–257 (2000).
80. X. Zhang, J. Xia, and K. Matyjaszewski, Macromolecules 32, 7349 (1999).
81. D. Benoit, E. Harth, P. Fox, R. M. Waymouth, and C. J. Hawker, Macromolecules 33,
363–370 (2000).
82. K. Matyjaszewski, D. A. Shipp, G. P. McMurthy, S. G. Gaynor, and T. Pakula, J. Polym.
Sci., Polym. Chem. Ed. 38, 2023 (2000).
83. K. Matyjaszewski, D. A. Shipp, J.-L. Wang, T. Grimaud, and T. E. Patten, Macro-
molecules 31, 6836 1998.
84. K. A. Davis and K. Matyjaszewski, Adv. Polym. Sci. 159, 2 (2002).
85. T. E. Patten and K. Matyjaszewski, Adv. Mater. 10, 901 (1998).
86. K. A. Davis and K. Matyjaszewski, Macromolecules 34, 2101 (2001).
87. K. Matyjaszewski, M. J. Ziegler, S. V. Arehart, D. Greszta and T. Pakula, J. Phys. Org.
Chem. 13, 775 (2000).
88. B. Kirci, J. F. Lutz and K. Matyjaszewski, Macromolecules 35, 2448 (2002).
89. C. J. Hawker, J. M. J. Frechet, R. B. Grubbs and J. Dao, J. Am. Chem. Soc. 117, 10763
(1995).
90. S. G. Gaynor, S. Edelman and K. Matyjaszewski, Macromolecules 29, 1079 (1996).
91. K. Matyjaszewski, S. G. Gaynor, and A. Kulfan, Macromolecules 30, 5192 (1997).
92. J.-S. Wang, D. Greszta and K. Matyjaszewski, Polym. Mater. Sci. Eng. 73, 416
(1995).
93. K. Matyjaszewski, P. J. Miller, E. Fossum and Y. Nakagawa, Appl. Organomet. Chem.
12, 667 (1998).
94. J. Xia, X. Zhang, and K. Matyjaszewski, Macromolecules 32, 4482 (1999).
95. S. G. Gaynor and K. Matyjaszewski, Macromolecules 30, 4241 (1997).
96. D. M. Haddleton, A. M. Heming, A. P. Jarvis, A. Khan, A. Marsh, S. Perrier, S. A. F.
Bon, S. G. Jackson, R. Edmonds, E. Kelly, D. Kukuli, and C. Waterson, Macromol.
Symp. 157, 201–208 (2000).
97. K. Ohno, B. Wong, and D. M. Haddleton, J. Polym. Sci., Part A: Polym. Chem. 39,
2206–2214 (2001).
98. X. Chen, B. Ivan, J. Kops, and W. Batsberg, Macromol. Rapid Commun. 19, 585–589
(1998).
99. U. M. Stehling, E. E. Malmstroem, R. M. Waymouth, and C. J. Hawker, Macro-
molecules 31, 4396–4398 (1998).
100. R. Jerome, D. Mecerreyes, D. Tian, P. Dubois, C. J. Hawker, M. Trollsas, and J. L.
Hedrick, Macromol. Symp. 132, 385–403 (1998).
101. K. Yamada, M. Miyazaki, K. Ohno, T. Fukuda, and M. Minoda, Macromolecules 32,
290–293 (1999).
102. D. Mecerreyes, B. Atthoff, K. A. Boduch, M. Trollsaas, and J. L. Hedrick, Macro-
molecules 32, 5175–5182 (1999).
103. F. Liu, S. Ying, N. Luo, B. Liu, Q. Liu, and Polym. Prepr. 40, 1032–1033 (1999).
104. X.-S. Wang, N. Luo, S.-K. Ying, and Q. Liu, Eur. Polym. J. 36, 149–156 (1999).
105. B. Liu, F. Liu, N. Luo, S.-K. Ying, and Q. Liu, Chin. J. Polym. Sci. 18, 39–43 (2000).
106. J.-D. Tong, S. Ni, and M. A. Winnik, Macromolecules 33, 1482–1486 (2000).
107. U. Meyer, A. R. A. Palmans, T. Loontjens, and A. Heise, Macromolecules 35, 2873–2875
(2002).
108. J. Qiu, B. Charleux, and K. Matyjaszewski, Prog. Polym. Sci. 26, 2083 (2001).
720
LIVING RADICAL POLYMERIZATION
Vol. 6
109. S. A. F. Bon, M. Bosveld, B. Klumperman, and A. L. German, Macromolecules 30,
324–326 (1997).
110. C. Farcet, B. Charleux, and R. Pirri, Macromol. Symp. 182, 249 (2002).
111. C. Farcet, J. Nicolas, and B. Charleux, J. Polym. Sci., Part A: Polym. Chem. 40, 4410–
4420 (2002).
112. S. Gaynor, J. Qiu, and K. Matyjaszewski, Macromolecules 31, 5951 (1998).
113. J. Qiu, S. Gaynor, and K. Matyjaszewski, Macromolecules 32, 2872 (1999).
114. K. Matyjaszewski, D. A. Shipp, J. Qiu, and S. G. Gaynor, Macromolecules 33, 2296
(2000).
115. J. Qiu, T. Pintauer, S. G. Gaynor, K. Matyjaszewski, B. Charleux, and J.-P. Vairon,
Macromolecules 33, 7310 (2000).
116. S. Jousset, J. Qiu, K. Matyjaszewski, and C. Granel, Macromolecules 34, 6641 (2001).
117. K. Matyjaszewski, J. Qiu, N. V. Tsarevsky, and B. Charleux, J. Polym. Sci., Part A:
Polym. Chem. 38, 4724 (2000).
118. I. Uzulina, S. Kanagasabapathy, and J. Claverie, Macromol. Symp. 150, 33–38 (2000).
119. M. J. Monteiro, M. Hodgson, and H. De Brouwer, J. Polym. Sci., Part A: Polym. Chem.
38, 3864–3874 (2000).
120. M. J. Monteiro, M. Sjoberg, J. Van der Vlist, and C. M. Gottgens, J. Polym. Sci., Part
A: Polym. Chem. 38, 4206–4217 (2000).
121. M. J. de Monteiro and J. Barbeyrac, Macromolecules 34, 4416–4423 (2001).
122. A. Butte, G. Storti, and M. Morbidelli, Macromolecules 33, 3485–3487 (2000).
123. J. G. Tsavalas, F. J. Schork, H. de Brouwer, and M. J. Monteiro, Macromolecules 34,
3938–3946 (2001).
124. A. Butte, G. Storti, and M. Morbidelli, Macromolecules 34, 5885–5896 (2001).
125. J. J. Vosloo, D. De Wet-Roos, M. P. Tonge, and R. D. Sanderson, Macromolecules 35,
4894–4902 (2002).
126. C. J. Ferguson, R. J. Hughes, B. T. T. Pham, B. S. Hawkett, R. G. Gilbert, A. K. Serelis,
and C. H. Such, Macromolecules 35, 9243–9245 (2002).
127. S. Viala, K. Tauer, M. Antonietti, R.-P. Kruger, and W. Bremser, Polymer 43, 7231–
7241 (2002).
128. C. Schilli, A. H. E. Mueller, E. Rizzardo, S. H. Thang, and B. Y. K. Chong, Polym.
Prepr. 43, 687–688 (2002).
129. S. Monge and D. M. Haddleton, Polym. Prepr. 43, 793–794 (2002).
130. K. S. Murthy, Q. Ma, C. G. Clark Jr., E. E. Remsen, and K. L. Wooley, Chem. Commun.
773–774 (2001).
131. J. Liu, Q. Zhang, E. E. Remsen, and K. L. Wooley, Biomacromolecules 2, 362–368
(2001).
132. T. von Werne and T. E. Patten, J. Am. Chem. Soc. 123, 7497–7505.
133. K. Ohno, K.-M. Koh, Y. Tsujii, and T. Fukuda, Macromolecules 35, 8989–8993 (2002).
134. S. Blomberg, S. Ostberg, E. Harth, A. W. Bosman, B. Van Horn, and C. J. Hawker, J.
Polym. Sci, Part A: Polym. Chem. 40, 1309–1320 (2002).
B
ERT
K
LUMPERMAN
Eindhoven University
LLDPE.
See E
THYLENE
P
OLYMERS
, LLDPE.
LOW DENSITY POLYETHYLENE.
See E
THYLENE
P
OLYMERS
, LDPE.