Polymerization-Induced Self-Assembly

After quite a challenging start, reversible-deactivation radical polymerization (RDRP) has been successfully implemented in dispersed media, and one of the most significant achievements is undoubtedly the development of synthetic routes allowing the production of block copolymer nano-objects. This approach, known as polymerization-induced self-assembly (PISA), takes advantage of the chain-end reactivity of solvophilic macromolecules obtained by RDRP for the polymerization of a second monomer in a suitable solvent. The growth of the second block, insoluble in the polymerization medium, leads to the formation of block copolymers that self-assemble into nanoparticles. PISA can be performed in emulsion polymerization conditions, in which the monomers are insoluble, or in dispersion polymerization conditions, in which the monomers are soluble. Under optimized conditions, PISA can directly produce the same self-assembled morphologies (spheres, rods, fibers, vesicles) previously obtained by the solvent-displacement method using preformed block copolymers, but at much higher solids contents (up to 40–50%) and with significantly less experimental effort. The great majority of the PISA systems reported so far rely on dispersion polymerization, in which particle morphology is generally more easily tuned. The concept was originally developed by the team of Robert Gilbert and Brian Hawkett in Sydney in 2002,1 even if the name "polymerization-induced self-assembly" was not used at that time. PISA has become very popular over the last decade (Figure 1). The success of PISA can be explained by the ease and versatility of the method (applicable to almost any kind of monomers, polymerized in organic solvents or aqueous media), and of course by the wide variety of systems it gives access to, especially in terms of particle composition, morphology, and functionality. So far, the most versatile (in terms of monomer and solvent compatibility) and reliable polymerization technique for PISA has proved to be the reversible addition-fragmentation chain transfer (RAFT) polymerization technique. Other RDRP techniques have also successfully been implemented in the past, such as atom transfer radical polymerization (ATRP), nitroxide-mediated polymerization (NMP), organotellurium-mediated radical polymerization (TERP), organometallic-mediated radical polymerization (OMRP), and iodine transfer polymerization (ITP). These techniques, however, present various drawbacks compared to RAFT polymerization, and remain much less exploited in the field of PISA. Very recently, it was shown that PISA is not only feasible with RDRP, but that the principles also hold for other polymerization mechanisms, in particular ring-opening metathesis polymerization (ROMP). The recent trends in RDRP towards photo-initiated mechanisms have also been transposed to heterogeneous polymerization systems including photo-induced PISA. This special issue on PISA, which presents 23 papers spanning the range of current research in this growing field, is representative of the aforementioned trends. Most of the articles make use of RAFT polymerization (see below) but other techniques are also explored. For instance, two articles report ATRP-mediated PISA, namely ICAR ATRP2 and ARGET ATRP.3 Interestingly, the former contribution2 also compares ICAR ATRP-PISA to a standard RAFT-PISA system. Zhang, Cheng, Zhu, and co-workers4 used light to develop an efficient ITP-based PISA system relying on in situ bromine-iodine transformation. The potential of light to control PISA is also demonstrated in the works of Tan et al.5 and O'Reilly et al.6 to photo-control RAFT-mediated dispersion polymerizations. Interestingly, the articles by Zhang, Cheng, Zhu et al.4 and by Matyjaszewski, Pietrasik et al.7 report the synthesis of the same copolymers in a dispersion polymerization process using two different polymerization techniques: ITP-photo-PISA in methanol, and thermally-initiated RAFT in ethanol, respectively. While completely different molar masses were targeted, both articles clearly highlight the importance of the monomer concentration on the kinetics, in particular on the nucleation, and on the resulting particle morphology. In addition, Matyjaszewski, Pietrasik et al.7 demonstrate that post-polymerization annealing of the particle dispersions may induce morphological changes. An alternative way to tune particle morphology is proposed by Zetterlund and co-workers by conducting PISA under different CO2 pressures using a CO2-responsive macroRAFT agent stabilizer.8 The group of Rieger and Stoffelbach9 used standard experimental conditions, but tuned the particle morphology by using multifunctional RAFT agents generating amphiphilic triblock and star copolymers. They rationally correlated the observed particle morphology with the macromolecular characteristics. Whereas the great majority of the particles were synthesized via RDRP techniques, implementation of ROMP for PISA in water is also reported, either in dispersion10 or in emulsion.11 Stable diblock copolymer particle dispersions could be successfully synthesized using both polymerization processes, but the formation of anisotropic morphologies or vesicles was not observed. It is well established that the dispersion polymerization mechanism, where the monomer is soluble in the polymerization medium, favours the control over the polymerization and the formation of higher-order morphologies, and is therefore usually preferred over emulsion polymerization. Similarly, the majority of contributions report on dispersion polymerization systems—mostly performed in alcohols, water, or their mixtures—and generally a large variety of morphologies were generated, whereas the rarer examples performed in emulsion11-16 yielded spherical particles apart from the work of Hawkett et al.,15 where nanofibers were formed. The presence of a crosslinker during PISA can also prevent nanoparticle reorganization and inhibit the formation of higher-order morphologies. As such, in the work of Wang, Matyjaszewski, and co-workers,3 reporting on the simultaneous ARGET ATRP PISA (in dispersion) and the in-situ crosslinking through ring opening of glycidyl methacrylate (GMA), only spherical particles were obtained. Two additional contributions propose alternative strategies to crosslink the core of the block copolymer particles: O'Reilly et al.6 propose the polymerization of pentafluorophenyl methacrylate (PFMA) as a core monomer yielding spherical particles, which could be successfully crosslinked after polymerization by the addition of diamines in a dry solvent (DMSO) to avoid side reactions. Alternatively, Roth et al.17 report the dispersion polymerization of pentafluorobenzyl methacrylate in ethanol, which appears more stable than PPFMA to side reactions and very reactive towards thio-para-fluorosubstitutions. The reactivity of the particle core was studied with mono- or dithiols, the latter leading to core-crosslinked particles. In addition to core-crosslinked particles, the possibility to produce shell-crosslinked spheres is also reported: Tan et al.5 chose diacetone acrylamide as a shell constituting monomer, and after particle formation the shell of various particle morphologies was crosslinked via a ketone–amine reaction. Several other papers in this special issue focus also on the synthesis of reactive and/or functional particles. Whereas in the previous articles the shell or the core block is generally a homopolymer of a reactive monomer, the contribution from Lovett, Armes et al.12 aimed at inserting one reactive epoxy group per chain on average in the shell of the nanospheres, either by an epoxy-functional RAFT agent or through a stepwise copolymerization strategy using GMA as a comonomer. They demonstrated that the epoxy function was much more resistant towards hydrolysis when it was inserted within the polymer chain, than when it was exposed at the α-chain end, and that one single reactive group on average was sufficient to post-functionalize the particles by reaction with primary amines. Aiming at the synthesis of building blocks for larger supramolecular structures, Lansalot, D'Agosto et al.14 reported the possibility to functionalize the surface of polystyrene particles obtained in emulsion polymerization by naphthalene units present again at the α-chain end of the macroRAFT agents. They overcame the challenge to localize an intrinsically hydrophobic moiety at the surface of particles dispersed in water, by using charged polyacrylic acid (PAA) as a stabilizer. Three articles propose the synthesis of stimuli-responsive particles. While Gianneschi et al. inserted peptide sequences in the particle shell that could be cleaved via enzymes inducing particle degradation/aggregation,10 the teams of Boyer18 and Lacombe and Save13 designed photo-active particles. Boyer et al. propose the insertion of 1-pyrenemethyl methacrylate as a comonomer in the solvophobic block, as a means to photochemically trigger particle dissociation by cleavage of pyrene moieties leading to a solvophobic–solvophilic transition of the core polymer. Save and Lacombe not only incorporated a photo-reactive monomer in particle shell (i.e., 2-Rose Bengal ethyl acrylate), but they also proposed their use as film-forming particles to produce photoactive films exhibiting photo-oxidation activity. Completely different applications are targeted in the following two articles that propose the synthesis of magneto-responsive composite materials. Hawkett et al.15 designed superparamagnetic iron oxide nanoparticles (SPION)-decorated nanofibers by electrostatic complexation of SPION and nanofibers, post-polymerization. The resulting composites displayed magnetic properties and aligned under a magnetic field. In contrast, Semsarilar et al.19 performed PISA in the presence of oleic-acid-stabilized SPION and successfully obtained the desired composites of various morphologies. They were used to construct magneto-responsive membranes, of which the hydraulic permeability could be modulated via magnetic field through structural rearrangement of the nanoparticles leading to changes in the void space. Apart from these particular applications, the most frequently considered uses of block copolymer particles prepared by PISA in water are related to the biomedical field. Perrier et al.16 explored the potential of spherical polystyrene nanoparticles surrounded by a poly(2-acrylamido-2-methylpropane sodium sulfonate) (PAMPS) shell to mimic the activity of heparin. They notably showed that these particles out-performed heparin and linear PAMPS in cellular proliferation, while keeping a low toxicity. Such low cytotoxicity was also briefly demonstrated for the above-mentioned particles prepared by the O'Reilly's group.6 The potential and limits of PISA-derived particles for biomedical applications are in fact well assessed in the review article of Hong, Pan et al.20 (focusing on drug delivery) and in the feature article of Truong, Davis et al.,21 whereas Cheng and Perez-Mercader22 give a prospective view on the potential of PISA in the field of artificial biology. Apart from these three biomedical application-oriented review articles, the feature article of An and Wang23 gives a more general overview of RAFT-mediated PISA in dispersion, highlighting the great efforts already made to identify the various factors affecting particle morphology and stability. Finally, Delaittre et al.'s review article focuses on the introduction of reactivity and functionality to PISA-derived nano-objects.24 These review and feature articles provide some interesting perspectives for PISA. To date, the applications of PISA have been mainly centred on the biomedical field, where most efforts seem to be focused at the moment. The examples in this special issue, however, demonstrate possible applications in other fields (e.g., smart coatings and membranes). We should thus not lose sight of the great potential of PISA for the synthesis of various functional objects (including hybrids) that could find applications in various fields, such as catalysis, coatings, or additives (e.g., stabilizers, viscosifiers, opacifiers, etc.). RAFT polymerization remains the predominant and most versatile polymerization technique, and the fact that the pioneering RAFT polymerization patents have passed into the public domain in 2018 should promote industrial transposal of PISA. NMP has also been successfully used for PISA, even if no papers report on that technique in this issue. The recent successful implementation of other polymerization techniques such as ROMP will undoubtedly broaden the chemical nature of nano-objects, paving the way towards new applications of PISA-derived particles. We are very excited to bring you this special issue on PISA. We hope you enjoy it and that it will stimulate new ideas for your own research and initiate fruitful collaborations.