CONTRIBUTION OF CONTROLLED RADICAL POLYMERIZATION TO THE
2.2 PRELIMINARY COMMENTS UNDERLYING CONTROLLED RADICAL POLYMERIZATION
When discussing the origins of the PISA process, it is important to recognize that it is not restricted to controlled radical polymerization techniques, and that any living polymerization technique that allows the reactivation of a soluble polymer with the formation of a second, insoluble polymer block could be referred to as a PISA pro- cess. Work published almost 40 years ago by Barrett [15], and more recently by Okay et al. [22] and Kimet al. [23], exploited the livingness of poly(t-butyl styryl)-lithium chains to initiate the dispersion polymerization of styrene (St) [15, 23] or divinylben- zene [22] inn-hexane, and indeed pioneered the concept of what is today called a PISA process. PISA can, in essence, be defined as the self-assembly of (amphiphilic) block copolymers during polymerization. However, as it will be detailed below, some of the soluble living polymers used for the polymerization in dispersed media (emul- sion or dispersion) are in some cases amphiphilic and able to self-assemble before the polymerization starts [18, 24]. In both cases, however, particle formation relies on an efficient micellar nucleation. Both types of living precursors will thus be described in this chapter.
Since controlled radical polymerization techniques are central to the PISA process, they will be briefly described in the following sections.
2.2.1 Introduction
Free-radical polymerization is the most important method for the production of synthetic polymers in large-scale industrial production and in the manifold appli- cations. Despite the wide range of functional monomers that can be polymerized by a radical mechanism and the great variety of statistical copolymers with many structures and properties [25] that can be produced by free-radical polymerization,
k k
NMP + O–N
X+ MetXn/Ligand
Y Y
+
+
MetXn+1/Ligand
+ R2 R1 ATRP
DT or RAFT
O–N R2 R1
Figure 2.3 Commonly used techniques to control the free radical polymerization.
Nitroxide-mediated polymerization (NMP), atom transfer radical polymerization (ATRP, Met
= metal atom, X = halogen atom), degenerative-transfer polymerization (DT, −Y= −I,
−TeCH3) or reversible addition–fragmentation chain transfer polymerization (RAFT,−Y=
−SC(=S)−Z ; Z=activating group).
the unavoidable occurrence of fast radical–radical termination reactions leads to a lack of control over the resulting polymers’ molar masses, molar mass distributions, chain-end functionalities and the macromolecular architecture. The pathway found to get around this is the emergence of the so-called living or controlled radical polymerization techniques (IUPAC: reversible-deactivation radical polymerization, RDRP) that has opened a new era for the free-radical polymerization. The CRP techniques can be categorically subdivided into two groups. They are either based on a reversible termination reaction or on a reversible chain transfer reaction (Figure 2.3). In both cases macromolecular radicals, called “active species,” undergo reversible deactivation—that is, successive activation–deactivation cycles—by being capped by a specific group under a form called dormant. Only a very small fraction of chains are simultaneously active and can propagate, but as a consequence of the rapid exchange of the radical between active and dormant macromolecular chains, they grow simultaneously during the whole polymerization period. The main feature of CRP is that the number-average molar masses increase linearly with monomer conversion and the corresponding molar mass distributions are narrow. This requires that a fast exchange between active and dormant chains takes place. After the monomer has been consumed, the formed chains still bear the mediating function at the chain end and can thus be further extended with either the same or another monomer in a subsequent polymerization step. This paves the way for the synthesis of block copolymers and other more complex architectures.
2.2.2 Major Methods Based on a Reversible Termination Mechanism
Nitroxide-mediated polymerization [26] and atom transfer radical polymerization [27] represent the two most commonly used methods of controlled radical polymer- ization based on a reversible termination reaction. In these polymerization systems, the dormant species are either an alkoxyamine for NMP or an alkyl halide for ATRP (Figure 2.3). In NMP, the homolytic cleavage of the alkoxyamine is thermally activated, which generally requires elevated temperatures, whereas in ATRP, the
k k activation of the alkyl halide chain ends is achieved through a redox reaction with
a transition metal complex under a broader range of conditions. In both cases the polymerization kinetics are governed by the activation–deactivation equilibrium and by the persistent radical effect [28]. The theoretical DPn can be calculated by the ratio of the initial monomer concentration and the initiator (i.e., alkoxyamine or alkyl halide) concentration, multiplied by the monomer conversion.
2.2.2.1 Nitroxide-Mediated Controlled Radical Polymerization (NMP) Nitrox- ides are stable radicals that are able to trap carbon-centered radicals at a nearly diffusion-controlled rate. While at low temperatures the formation of the alkoxyamine is kinetically irreversible, the formed C–O bond may undergo homolytic cleavage at elevated temperatures, regenerating a propagating radical and the nitroxide. This equilibrium between propagating radicals and dormant macroalkoxyamines governs the NMP. The initiation step can be performed in two different ways: either in a bicomponent system using a classical rad- ical initiator in combination with the free nitroxide or in a monocomponent system based on a preformed alkoxyamine. The latter is nowadays the most popular one because it allows the kinetics and the molar masses of the obtained polymer chains to be tuned very precisely. The typically used nitroxides are 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO), which was successfully applied for St and derivatives, and N-tert-butyl-N-(1-diethyl phosphono-2,2-dimethyl propyl) nitroxide (SG1) and 2,2,5-trimethyl-4-phenyl-3-azahexane-3-oxyl (TIPNO) (Figure 2.4), which have significantly expanded the range of monomers amenable for use in NMP and reduced the required polymerization times and temperatures.
2.2.2.2 Atom-Transfer Radical Polymerization (ATRP) ATRP [27] is based on the reversible transfer of a halogen atom between a dormant alkyl halide and a tran- sition metal catalyst in a redox reaction (Figure 2.3). The alkyl halide is reduced to a growing radical while the transition metal is oxidized. In most of the studied reac- tions, Cu(I) species complexed by multidentate amine ligands constitute the activator and the corresponding Cu(II) species are the deactivator. The initiation step can be performed in different ways, starting either from Cu(I) or from Cu(II). With Cu(I), the process is called direct ATRP. An alkyl halide initiator is employed. With Cu(II), there are multiple initiation possibilities, including the reverse ATRP process [27],
Figure 2.4 Chemical structures of commonly used nitroxides in NMP.
k k which uses a classical radical initiator; the simultaneous reverse and normal initiation
(SR&NI) method [29], which combines an alkyl halide along with a small fraction of a classical radical initiator (the latter produces radicals able to reduce Cu(II) to Cu(I)); the activator (re)generated by electron transfer (A(R)GET) technique [30], which uses an alkyl-halide initiator in combination with a reducing agent to turn Cu(II) into Cu(I); and e-ATRP [31], which utilizes the concept of ARGET-ATRP via an electrochemical stimulus to provide enhanced polymerization control. ATRP can be successfully applied to a broad variety of monomers using a multitude of avail- able mono- or multifunctional initiators and catalysts that allow the polymerization and polymer characteristics to be fine-tuned.
2.2.3 Major Methods Based on a Reversible Transfer Mechanism
For the methods proceeding according to a reversible transfer mechanism, a bimolec- ular reaction between an active and a dormant macromolecule leads to the transfer of the functional end-group (Figure 2.3). This can be a direct exchange as in the so-called degenerative transfer (DT) technique where an iodine atom is interchanged ((reverse) iodine transfer polymerization, (R)ITP) [32–34]. The organo-tellurium mediated CRP (TERP) follows a similar mechanism [35], with the exchange of a terminal –TeCH3 group. This reaction proceeds via both thermal dissociation and degenerative transfer. However, when an external source of free radicals is used at low temperatures, only degenerative transfer takes place. Another technique is the well-known reversible addition–fragmentation chain transfer (RAFT) [36, 37]
polymerization in which chains are end-functionalized by an unsaturated group (typically the C=S bond from a dithioester, a dithiocarbonate, a dithiocarbamate, or a trithiocarbonate). The radical is transferred through an addition–fragmentation step with a tertiary carbon-based radical as the intermediate. A reversible transfer system requires the use of a conventional radical source in addition to a reversible chain transfer agent RY (Figure 2.3). The first step can be considered as a con- ventional transfer reaction to RY, creating new chains with a R group at one end and a Y atom or functional group at the other end. The next step is the transfer of Y from an end-functionalized chain to a propagating macroradical, which is a thermodynamically neutral (i.e., degenerative) process. This second step does not create new chains but contributes to the extension and continuous redistribution of the active function amongst the existing chains. If the initial concentration of the radical initiator is low with respect to the initial concentration of the transfer agent, a large majority of the macromolecules have the same RAFT agent-derived end-groups. The concentration of growing macromolecular chains becomes constant and close to the initial concentration of the RAFT agent, once the latter has been completely consumed. Only at that stage, and when the reversible transfer takes place, can a linear increase of the number-average molar mass with monomer conversion be observed (the DPn can be approximated by the ratio of the initial monomer concentration divided by the chain transfer agent concentration, multiplied by the monomer conversion). The main requirement is that the rate constant of the transfer reaction to RY should be large, which is the case for the RAFT process.