A Perspective on the History and Current Opportunities of Aqueous RAFT Polymerization

Abstract

Reversible addition-fragmentation chain transfer (RAFT) polymerization has proven itself as a powerful polymerization technique affording facile control of molecular weight, molecular weight distribution, architecture, and chain end groups - while maintaining a high level of tolerance for solvent and monomer functional groups. RAFT is highly suited to water as a polymerization solvent, with aqueous RAFT now utilized for applications such as controlled synthesis of ultra-high molecular weight polymers, polymerization induced self-assembly, and biocompatible polymerizations, among others. Water as a solvent represents a non-toxic, cheap, and environmentally friendly alternative to organic solvents traditionally utilized for polymerizations. This, coupled with the benefits of RAFT polymerization, makes for a powerful combination in polymer science. This perspective provides a historical account of the initial developments of aqueous RAFT polymerization at the University of Southern Mississippi from the McCormick Research Group, details practical considerations for conducting aqueous RAFT polymerizations, and highlights some of the recent advances aqueous RAFT polymerization can provide. Finally, some of the future opportunities that this versatile polymerization technique in an aqueous environment can offer are discussed, and it is anticipated that the aqueous RAFT polymerization field will continue to realize these, and other exciting opportunities into the future.

1. Introduction

In the introduction of this perspective Prof. Charles L. McCormick recounts the early days of aqueous RAFT polymerization within his laboratory at the University of Southern Mississippi. The impetus for the development of aqueous RAFT polymerization techniques in our labs arose from a compelling need to prepare well-defined, stimuli-responsive polymers and copolymers from functional monomers directly in water under controlled, homogeneous reaction conditions. Although many conventional polymerization procedures were available, targeted architectures often required complicated reaction schemes with organic solvents, protected monomers, and postpolymerization modification. The original Commonwealth Scientific and Industrial Research Organization (CSIRO) communication on RAFT in December of 1998^[1] caught our immediate attention, and we began to explore the potential of RAFT polymerization of a wide variety of monomers in water that had been utilized for over two decades in our research involving controlled-release, enhanced oil recovery, drag reduction, and bioconjugation.

Our first attempts at aqueous RAFT polymerization of several neutral and ionic acrylamido monomers utilizing dithiobenzoates were quite disappointing, especially to graduate students in our labs who were ready to move on from difficulties and limitations experienced with other current “living/controlled” techniques such as nitroxide mediated and atom transfer radical polymerization (NMP and ATRP, respectively). Slow monomer conversion and no control over the molecular weight or the molecular weight distribution were observed in the presence of these initial dithiobenzoates. Control experiments in the absence of RAFT chain transfer agents (CTAs) proceeded at the expected rates and yielded high molecular weights and broad molecular weight distributions. It was clear that RAFT polymerization in water was not going to be trivial! At our next group meeting, the consensus was that CTA and dormant dithioester chain ends were likely hydrolyzed under our aqueous conditions. A quick experiment over the weekend by Mitsukami and Donovan in which the pH was lowered to 5–6 demonstrated some monomer conversion to polymer, but again with little control. Indeed, later detailed kinetic experiments by Mitsukami, Donovan, Sumerlin, and Thomas involving hydrolysis of dithiobenzoates and a trithiocarbonate, clearly suggested that lowering both the pH and temperature would help mitigate competitive loss of CTA and dormant end groups (discussed in detail below). In retrospect, lowering temperature and pH and choosing trithiocarbonate RAFT agents for aqueous polymerizations were quite fortuitous.

Statistical and block copolymerizations conducted under proper conditions with appropriate CTAs yielded target molecular weights and narrow molecular weight distributions as confirmed by size exclusion chromatography multiangle laser light scattering (SEC/MALLS). Unlike our initially-chosen dithiobenzoates, which are less hydrolytically stable, produce very stable intermediate radicals, and show significant retardation effects—trithiocarbonates are stable to both hydrolysis and aminolysis, have short-lived, high potential energy intermediates, and are less prone to retardation. Fortunately, with trithiocarbonates, we were also spared the long “initialization” periods elucidated so well by Klumperman et al.,^[2] which plagued our first experiments. Of course, it is still necessary to choose Z and R groups that are monomer-appropriate. It has been essential to synthesize RAFT agents in many cases, and a large body of open and patented literature exists regarding those synthetic methods. As well, our group found that choosing water-soluble initiators with appropriate half-lives (yielding sufficient radical flux) is also important, especially for ambient temperature RAFT.^[3] This was demonstrated in a study led by Convertine where the direct synthesis of thermally responsive block copolymers based on N-isopropylacrylamide (NIPAM) in aqueous media at room temperature was achieved.^[4]

Throughout the years many great students and post-docs worked on aqueous RAFT polymerization within the McCormick Research Group, with a complete list of the graduate student contributors provided in the Supporting Information. Within a decade of the first reports, many anionic, cationic, neutral, and zwitterionic monomers were directly polymerized under aqueous conditions by RAFT, resulting in well-defined architectures. In this perspective, we examine not only the early studies of aqueous RAFT but also recent advances providing details on the practicalities of aqueous RAFT polymerization, current applications, and future opportunities of this versatile yet powerful technique.

2. Practical Considerations of Aqueous RAFT Polymerization

Since its’ introduction in 1998, RAFT polymerization has grown to be one of the most powerful techniques for preparing complex polymer architectures with low molar mass dispersity, predictable molecular weights, and high chain end-group fidelity.^[1] One of the many benefits of RAFT polymerization is that it can be conducted in an aqueous solvent. Aqueous solvents allow for faster polymerization kinetics as well as a green and biocompatible reaction medium. Using water as the reaction solvent opens the door for RAFT polymerization in processes like bioconjugation, and biocompatible polymerizations, all without the need for toxic solvents. However, an aqueous solvent also introduces other potential constraints to the polymerization process, such as hydrolysis and aminolysis of the chain transfer agents. Hence careful consideration of polymerization conditions is required for optimum control of the polymerization process. Herein we will provide some practical considerations for successfully conducting aqueous RAFT polymerization.

2.1. Chain Transfer Agents (CTAs)

RAFT polymerization utilizes chain transfer agents (CTAs) in the form of thiocarbonylthio compounds that can provide control of the polymerization by mediating it through a reversible chain-transfer process. CTAs can be divided into four main categories: dithioesters, trithiocarbonates, xanthates, and dithiocarbamates (Figure 1). In general, dithioesters and trithiocarbonates are more suitable for controlling polymerization of the “more activated monomers” (MAMs) like methacrylate, methacrylamide, and styryl derivatives while xanthates and dithiocarbamates are more suitable for “less activated monomers” like vinyl esters and vinyl amides.^[5] Since the most common types of monomers encountered in RAFT polymerization in general (and aqueous RAFT in particular) are MAMs, the majority of aqueous RAFT literature, and our focus herein when discussing CTAs, will be on dithioesters and trithiocarbonates.

Figure 1.

Common classes of CTAs that can be utilized to synthesize well-defined polymers and block copolymers using aqueous RAFT polymerization.

A wide variety of homopolymers and block copolymers have been prepared directly in aqueous media utilizing water-soluble CTAs. These water-soluble CTAs can be either small-molecule CTAs, or they can be formed from water-soluble thiocarbonylthio-functionalized polymers (macroCTAs) that can be chain extended with additional monomer to prepare block copolymers. Since many small molecule CTAs exhibit limited water solubility, employing a macroCTA that is comprised of a hydrophilic polymer may be advantageous in RAFT polymerization by increasing the solubility of the CTA. However, before conducting an aqueous RAFT polymerization, careful attention must be paid to mitigate certain competing reactions such as hydrolysis and aminolysis of the CTA which can negatively impact the ability of the CTA to provide control over the polymerization. A few strategies that can mitigate these reactions, are briefly discussed below.

2.1.1. Hydrolysis and Aminolysis of CTAs

Thiocarbonylthio compounds are susceptible to hydrolysis since they are sulfur analogs of esters.^[6] This can lead to uncontrolled polymerizations resulting in polymer chains with high dispersity and deviations from predetermined molecular weights. In 2000, Levesque et al. investigated the effect of pH and temperature on the hydrolytic stability of thiocarbonylthio compounds in water.^[6a] By varying the temperature from 20 to 35 °C and the pH from 7.5 to 8.5, they found that the hydrolysis rate of the thiocarbonylthio compounds increased with both increasing temperature and pH. The McCormick group performed some of the first examinations of the effect of temperature and pH on the hydrolysis of CTAs, including a dithiobenzoate CTA, sodium 4-cyanopentanoic acid dithiobenzoate (CTP), and two CTP-based poly(sodium 2-acrylamido-2-methylpropanesulfonate) (AMPS) macroCTAs with degrees of polymerization of 9 and 38, respectively.^[6b] In analyzing the pseudo first-order rate plots for the hydrolysis of CTP and the two CTP-based macroCTAs (CTP-AMPS₉ and CTP-AMPS₃₈), it was demonstrated that the rate of CTP hydrolysis increased with increasing pH and that the rate of hydrolysis of the two macroCTAs decreased with an increasing molecular weight of the polymer chains (Figure 2).^[6b] They attributed the effect of decreasing hydrolysis rate with an increase of macroCTA molecular weight to the polymer chains’ ability to protect the dithioester moiety from the attack of water molecules through sterics.^[6b]

Figure 2.

a) Molecular structure of CTP and CTP-AMPS_n macro-CTAs and pseudo first-order rate plots for the hydrolysis of b) CTP, c) CTP-AMPS₃₈, and d) CTP-AMPS₉ at 70 °C. Reprinted with permission.^[6b] Copyright 2004, American Chemical Society.

The McCormick group followed this work by investigating the effect of temperature on the rate of hydrolysis of a trithiocarbonate-based CTA 2-(1-carboxy-1-methylethylsulfanylthiocarbonylsulfanyl)-2-methyl-propionic acid (CMP) under acidic conditions at pH 5.5 (Figure 3a).^[6c] Remarkably, at temperatures below 50 °C, the rate of CTA hydrolysis was negligible. Cai and co-workers also investigated the stability of another type of trithiocarbonate CTA, S-1-Ethyl-S′-(α,α′-dimethyl-α″-acetic acid)trithiocarbonate (EDMAT) at different pH values under mild light irradiation and ambient temperatures (Figure 3b).^[7] At highly acidic pH (pH = 2.6), there was no discernable hydrolysis of the CTA over a period of 4 h of irradiation. In contrast, hydrolysis was detected at neutral pH and increased in alkali solutions. It is necessary to emphasize that the CTAs studied here were trithiocarbonates which are considered more hydrolytically stable than their dithiobenzoate counterparts in aqueous media.^{[6b, 6c, 7,8]}

Figure 3.

a) Fraction of CTA remaining as a function of time at pH 5.5. Reproduced with permission.^[6c] Copyright 2005, Wiley-VCH. b) Spectroscopic stability measurements of EDMAT in aqueous solutions of varying pH values. A₀ is the solution absorbance before irradiation, and A is the solution absorbance at predetermined time intervals. Reproduced with permission.^[7] Copyright 2009, American Chemical Society.

CTAs are also susceptible to aminolysis by primary and secondary amines and monomers that possess these moieties. As a result, such monomers were initially thought to be precluded from direct RAFT polymerization. Additionally, certain amine-containing monomers can undergo hydrolysis and produce primary and secondary amines, which can react detrimentally with the CTAs. Studies; however, soon demonstrated that lowering the pH of a solution containing amines and thiocarbonylthio compounds can protonate the amine-containing compounds and significantly reduce aminolysis.^{[6a, 9]} In their report mentioned above, investigating the hydrolysis of CTP and CTP-based macroCTAs, the McCormick group also examined the effect of aminolysis on CTP and AMPS-based CTP macroCTAs by exposing them to ammonia in buffered aqueous solutions at pH values of 5.5 and 7 where both hydrolysis and aminolysis may play a role in the degradation of the CTA.^[6b] This work demonstrated the CTA concentration decreased more rapidly at pH 7 compared to pH 5.5, which the authors attributed to the dual role of hydrolysis and aminolysis at the higher pH since the amines were protonated at lower pH. Additionally, the concentration of the small-molecule CTA decreased more rapidly than the macroCTA, presumably due again to steric shielding of the CTA moieties from ammonia attack. Utilizing acidic conditions to control CTA degradation allowed for the successful and controlled aqueous RAFT polymerization of the primary amine-containing monomer acrylamide.^{[6b, 10]} They subsequently demonstrated the ability to polymerize primary amine-containing monomers in an aqueous RAFT polymerization by lowering the solution pH.^[11] In both cases, they reported good control of the molecular weight and low polydispersity.

Thus, to minimize the effects of hydrolysis and/or aminolysis on the CTAs during an aqueous RAFT polymerization, certain key reaction conditions should be optimized, including lowering temperature; lowering the pH; utilizing macroCTAs; utilizing trithiocarbonates instead of dithiobenzoates when possible; and minimizing overall reaction time where possible. To achieve low temperatures and short polymerization times, which depend on the source of initiation used for the RAFT polymerization, several different techniques have been utilized to optimize these parameters in aqueous systems.

2.2. Initiation Systems

Since RAFT polymerization is a controlled radical polymerization conducted in the presence of a CTA, initiation systems commonly employed in free radical polymerization can also be used in aqueous RAFT polymerization as long as the initiators are amenable to producing radicals in water. Some of these systems include: thermal initiators;^{[6c, 12]} photoradical initiators;^[13] photocatalysts;^[14] redox initiators;^[15] photoiniferter;^[16] and sound waves.^[17] Two of the most common and versatile initiation systems include thermal and photoinitiation methods of which we will discuss in detail below.

2.2.1. Thermal Initiators

In early reports of aqueous RAFT polymerization, thermal initiation via the cleavage of azo-containing species at elevated temperatures was the predominant initiation method.^[1,12,18] Certain common azo-initiators, like 2,2′-Azobis((2-methylpropionitrile) (AIBN), are sparingly water-soluble, but some water-soluble examples used in aqueous RAFT are 4,4′-azobis(4-cyanopentanoic acid) (V-501),^{[18a, 18b, 18e]} and 2,2′-azobis[2-methylN-(2-hydroxyethyl)propionamide] (VA-086).^[18c] As mentioned previously, reaction temperature can influence the rate of hydrolysis of CTAs, so utilizing thermal initiators with low activation temperatures may be beneficial when performing aqueous RAFT, especially at increased reaction times. It is also worth noting that oxygen must be removed before polymerization when using thermal initiators in a RAFT process.

2.2.2. Photopolymerization

Light-mediated aqueous RAFT photopolymerization can be divided into three main categories: externally catalyzed, catalyst-free (photoiniferter), and photoinduced electron/energy transfer (PET)-RAFT. In addition to thermal methods, RAFT polymerization can be initiated photochemically via direct photoactivation of CTAs by ultraviolet (UV) and in some cases visible light. UV light can cause photolysis of CTA moieties that can lead to R group fragmentation utilized in photoiniferter polymerization (discussed below),^[19] but can also result in CTA degradation and side reactions resulting from the high energy activation. While direct photoactivation of CTAs with visible light has been reported, the conditions to achieve this direct photoactivation are not trivial and highly dependent on light wavelengths, intensities, monomers, and specific CTA compounds, often requiring high intensity light to achieve polymerization.^[20] Significant work by Cai and co-workers has probed the conditions required to achieve direct RAFT photoiniferter polymerization under visible light.^[7,21] Visible light has garnered much interest as an initiation source in contrast to other methods since it can be achieved with low energy, relatively cheap and safe, light-emitting diodes (LEDs). The low-energy irradiation reduces the chance of damage to other biomolecules or sensitive reagents; can be conducted at ambient temperatures, which can reduce the rates of CTA hydrolysis; and is often cheap and readily available with some studies even utilizing direct sunlight.^[22] In contrast with UV irradiation, visible light-initiated systems may require longer reaction times, but this can often be overcome by increasing the irradiation intensity and/or employing fast propagating monomers like acrylamides. Another attractive quality of photo-RAFT polymerization is that it allows for spatiotemporal control of the polymerization process by simply turning the light source on or off.^{[7,13,22b,23]} A recent example of this was demonstrated by McClelland et al. where they utilized CdSe quantum dots as photocatalysts for aqueous photo-RAFT polymerization where the nanoparticle size of the photocatalyst allowed for facile separation and recycling of the catalyst from the polymer product (Figure 4).^[14b] In this work, when irradiated with visible green light, polymer conversion increases, and when it is off, the conversion is at a standstill (Figure 4).^[14b] As a result of this spatiotemporal control of polymer conversion, most reports of aqueous RAFT photopolymerization have been visible light-mediated.

Figure 4.

a) Quantum dot photoinitiated aqueous RAFT polymerization, demonstrating light source control of b) monomer conversion reaction kinetics and c) control of polymer molecular weight as evidenced by representative gel permeation chromatography (GPC) traces. Reproduced with permission.^[14b] Copyright 2020, American Chemical Society.

Aqueous RAFT with External Photoinitiators:

Aqueous RAFT polymerization can be initiated photochemically via external photoinitiators that absorb light, fragment to produce radicals, and in turn initiate the RAFT polymerization. The use of photoinitiators also provides an additional benefit for RAFT polymerization by allowing the reactions to proceed at room temperature, which should improve the hydrolytic stability of the CTAs. Common photoinitiators for aqueous RAFT polymerization are sodium phenyl-2,4,6-trimethylbenzoyl phosphinate (SPTP) and (2,4,6-Trimethylbenzoyl)diphenylphosphine oxide (TPO) since they are water-soluble and degrade rapidly upon exposure to UV or visible light irradiation.^[13,23,24] If the half-life of the photoinitiator is shorter than that of a traditional thermal radical initiator, the rate of polymerization of an aqueous photo-RAFT polymerization can be significantly faster than an analogous thermal-polymerization. For instance, Tan and Zhang performed kinetic studies to compare the rate of aqueous RAFT polymerization from a water-soluble glycerol monomethacrylate (GMA) macroCTA of N-(2-Hydroxypropyl)methacrylamide (HPMA) by either a photo-initiated polymerization at 25 °C or the thermally initiated polymerization utilizing V-501 at 70 °C.^[23b] They found that the rate of polymerization was significantly faster for the photopolymerization relative to the thermal system despite the difference in reaction temperature, suggesting an even faster rate would occur at the matched elevated temperature.^[23b]

Catalyst Free (Photoiniferter) Polymerization:

Introduced by Otsu in 1982, an iniferter refers to a molecule that can act as an initiator, transfer agent, and termination agent.^[25] Many thiocarbonylthio compounds can initiate polymerization via a light-induced iniferter (photoiniferter) mechanism since they absorb visible light due to the n → π* transition of the C=S bond.^[19a] When using a CTA as a photoiniferter, initiation occurs via homolytic cleavage of the C-S bond to generate a carbon-centered radical (R group) that is capable of initiating polymerization and a thiocarbonylthio radical (Z group) that can deactivate growing chains via reversible termination.³³ The photoiniferter RAFT mechanism is attractive since it does not require any exogenous initiating species, and thus the polymer chains derived from these processes have homogeneous end groups. Similar to other photopolymerizations, the kinetics of an aqueous photoiniferter polymerization can be controlled by manipulating the intensity of the light source. For instance, Lewis et al. reported a decrease in reaction time to reach >85% conversion of acrylamide (AM) from 12 h to 11 min by increasing the LED power from 6 to 208 W at λ = 402 nm without a significant loss in the control of the polymerization.^[26] They reported that polymerizations under high irradiation intensities, where the temperatures are not controlled, can lead to solution temperatures of up to 80 °C due to both the exothermic nature of the polymerizations and the intense light irradiation. They noted some loss in control of the polymerization of dimethylacrylamide (DMA) at higher irradiation intensities which could be attributed to an increase in the hydrolysis rate of the CTA at higher temperatures.^[26] Therefore, it is critical to take appropriate measures to maintain moderate reaction temperatures in aqueous RAFT photopolymerization in order to minimize CTA hydrolysis. Since photoiniferter polymerization does not require exogenous initiators and generates radicals directly from the CTA, it has been viewed as a potential pathway toward sequence-controlled polymers via the single unit monomer insertion (SUMI) approach. Recently, Aerts et al. reported the SUMI of DMA into a trithiocarbonate utilizing an aqueous RAFT photoiniferter procedure.^[16a]

Aqueous PET-RAFT Polymerization:

Another type of RAFT photopolymerization that has gained popularity in recent years is photo-induced electron/energy transfer (PET) which utilizes photocatalysts instead of traditional external initiators. PET-RAFT catalysts can mediate RAFT polymerization through a redox reaction with the CTA to generate a thiocarbonylthio-derived radical that can then initiate polymerization.^[27] PET-RAFT polymerizations can occur via an oxidative catalytic pathway or a reductive catalytic pathway depending on the presence of a reducing agents such as ascorbic acid or a tertiary amine (Figure 5). Benefits of PET-RAFT include low catalyst loading (can be in the parts per million (ppm) range), ambient temperatures, oxygen tolerance, and inexpensive light sources (visible light via LED lights are the dominant light source reported) suitable for conducting the polymerization.^[27,28] Further, recent studies have demonstrated the utility of this technique with heterogenous nanoparticle based photocatalysts, which allows for the simple separation and reuse of the photocatalyst for multiple polymerizations through centrifugation, offering opportunities for improved sustainability of this technique.^{[14b, 29]} These attributes have made PET-RAFT an intriguing technique for biological macromolecules and a potential technique for precision polymer manufacturing in industry.

Figure 5.

Photo-induced electron/energy transfer RAFT polymerization using a photocatalyst via either a) an oxidative catalyst pathway or, b) a reductive catalytic pathway (tertiary amine (NR₃) as reducing agent). Reproduced with permission.^[32] Copyright 2018, American Chemical Society.

One of the most common PET-RAFT catalysts is the organic dye Eosin Y (EY), which absorbs visible blue and green light.^[30] In order to provide insight into the use of EY for PET-RAFT polymerization, the Sumerlin and Boyer groups through collaboration performed mechanistic studies of EY under blue and green light irradiation using both oxidative and reductive polymerization pathways.^[22a] They found that an oxidative mechanism during green light irradiation is the best way to prepare precise polymers with predictable molecular weights when utilizing EY as a catalyst since certain CTAs can absorb blue light and undergo photolysis to initiate polymerization via a competing photoiniferter mechanism. Additionally, they found that lowering the solution pH resulted in slower polymerization rates, which they attributed to the partial protonation/deactivation of tertiary amine cocatalysts.^[22a] Cocatalysts (usually tertiary amines or reducing agents like ascorbic acid) are often included in a PET-RAFT polymerization to increase polymerization rates, leading to shorter reaction times and allowing for photocatalyst regeneration. An added benefit of including a cocatalyst in a PET-RAFT polymerization is that under certain conditions, the polymerizations can proceed in the presence of oxygen, which is useful for both educational and potential industrial applications.^[31]

2.2.3. Sono-RAFT

Another intriguing initiation method for aqueous RAFT polymerization is via ultrasonic irradiation. This method was first reported by Qiao et al. in 2017 and utilizes ultrasonic waves to generate initiating hydroxyl radicals from solvent water molecules.^[17a] Thus, the water acts as both an initiator and solvent (inisolv), without the requirement for organic initiators or organic solvents, and can yield complete monomer conversion.^[17a] Utilizing this method, Qiao et al. have been able to prepare polymer nano-objects via ultrasound initiated RAFT polymerization-induced self-assembly (sono-RAFT-PISA) to realize nanomorphologies including spheres, worms, and vesicles.^{[17b, 33]}

3. Progress in the Field of Aqueous RAFT Polymerization

3.1. Polymerization-Induced Self-Assembly (PISA)

Considerable attention has been devoted to investigating the formation of nano-objects by the assembly of block copolymers in solution. Traditional self-assembly of block copolymer micelles involves postpolymerization processes whereby the polymers are diluted with excess solvent. Because of the multiple steps and use of high volumes of solvent (<1 wt% solids), the scalability of the preparation of block copolymer nano-objects is limited. In contrast, PISA, first reported by Gilbert and co-workers in 2002,^[34] can be utilized to form nano-objects with various morphologies in situ at high solids content (typically 10–40 wt%) via aqueous RAFT polymerization. In a typical process, a water-soluble macroCTA homopolymer is chain extended in water with a second monomer that gradually becomes insoluble as the chain grows, driving the in situ self-assembly of diblock copolymer nano-objects. PISA has been utilized to form “traditional” morphologies, such as spheres, worms, vesicles, and non-traditional morphologies like jellyfish and others.^[35] The formation of higher-order structures in PISA can depend upon several factors, including the length of the stabilizer block, the rate of polymerization of the solvophobic block, temperature, and the initiation method.^{[24c, 36]} The accurate reproduction of particular morphologies can be potentially challenging, and the construction of phase diagrams is usually required.

Traditional PISA typically requires a two-step polymerization method involving the synthesis of the solvophilic block followed by purification and then chain extension with the solvophobic block. To streamline this process, the Boyer Lab and colleagues utilized an aqueous RAFT polymerization process for the scalable flow synthesis of poly(dimethylacrylamide)-b-(poly(diaceteone-stat-dimethyl acrylamide)).^[37] In this work, a variety of solid contents and block lengths for the solvophilic and solvophobic blocks were achieved expanding the compositional space of worms, jellyfish, and vesicular nano-objects achieved using flow chemistry, without the need for intermittent purification (Figure 6).^[37]

Figure 6.

a) Schematic of the flow reactor setup when coupling two reactors to achieve block copolymers without intermittent purification. b) Representative TEM micrographs of pDMA-b-p(DAAm-co-DMA) nanoparticles in a one-step synthesis chain extending pDMA with DP200, DP400, and DP600 at 17.5 wt% solid content; b) w = branched worms, hbw = highly branched worms, v = vesicles. Reproduced with permission.^[37] Copyright 2019, American Chemical Society.

In follow up works, Xu et al. demonstrated a one-pot PISA approach via the formation of a gradient copolymer through a gradual injection of the core-forming monomer in the presence of the solvophilic monomer.^[35,38] This approach was utilized for the monomer pairs DMA and diacetone acrylamide (DAAm) to form worms and for oligo(ethylene glycol)methyl ether methacrylate (OEGMA) and HPMA to form higher-order morphologies like spheres, worms, vesicles, and jellyfish micelles.^[35] Due to the gradient nature of the copolymers, the DMA/DAAm system could form a pure worm phase without the addition of a cosolvent or secondary solvophilic monomer, and the OEGMA/HPMA system exhibited thermoresponsive behavior. Thus, this method greatly streamlines the PISA process by reducing the overall number of synthetic steps, allowing for the facile tuning of the copolymer’s composition by changing the solvophobic monomer’s injection rate. These qualities make it attractive for potential scale-up and industrial applications.

In 2018, Mellot et al. introduced a method to form dispersions of block copolymer fibers in water via PISA.^[39] In their work, they modified a trithiocarbonate CTA with a bis-urea sticker that acted as a template to direct the morphology of their PISA nano-objects toward the fiber morphology. This templated PISA method allowed for the tuning of the diameters of the fibers by changing the degree of polymerization of the hydrophobic block. This contrasts with conventional PISA, where the overall system determines the diameter of the worms/fibers. Their method allows for the reproducible production of the fiber morphology at various diameters over a large experimental window and could be intriguing for potential scale-up processes.

3.2. Bioapplications

Using water as a solvent opens the door for polymer bioconjugation utilizing biomolecules that are incompatible or sensitive to organic solvents. Aqueous RAFT polymerization has been used to prepare synthetic polymer-biomolecule conjugates through a “grafting from” approach from various biomolecules, including proteins and DNA.^[40] The grafting from strategy involves the CTA’s attachment to the molecule’s surface followed by polymerization to yield the bioconjugates.

Much of the aqueous RAFT bioconjugation work has been reported utilizing photoinitiation systems due to the benign effect that visible light has on biomolecules while also avoiding elevated temperatures (used in thermal-initiated reactions), which may denature biomolecules such as enzymes.^[41] This has made PET-RAFT an especially attractive polymerization technique for these reactions since it can be initiated by visible light and can utilize relatively nontoxic (and even some bio-derived) photocatalysts and cocatalysts such as vitamin B2 (riboflavin), riboflavin 5′-mononucleotide (FMN) and vitamin C (ascorbic acid), respectively.^[42]

DNA-polymer hybrid materials have gained great interest due to their potential use in drug delivery. In 2019, Lueckerath et al. utilized PET-RAFT to prepare DNA-polymer conjugates from a CTA-terminated DNA sequence using EY as the PET-RAFT catalyst and ascorbic acid as the cocatalyst.^[40b] They utilized various monomer classes (methacrylates, acrylates, and acrylamides) and were able to achieve high molecular weight DNA-polymer conjugates (>30 000 g mol⁻¹) that can self-assemble into nanostructures. Following this work, in 2020, Lueckerath et al. introduced a novel platform to form isotropic or anisotropic DNA-polymer nanostructures via PISA after enzyme degassing from single-stranded DNA by varying the length of the polymer block (Figure 7).^[43] Through this method, they were able to form complex DNA-polymer architectures such as micelles and worms, which streamlines the preparation of DNA polymer nanostructures. Aqueous RAFT has also recently been applied to produce an arm-first, star polymer conjugated to a DNA core for the delivery of DNA for intracellular DNA delivery.^[44] In the synthesis of these DNA-polymer hybrid materials, the use of aqueous biocompatible solvents was found to lower the dispersity while increasing the solid content of the polymer produced. Finally, in 2017 Hawker and colleagues demonstrated the utility of aqueous PET-RAFT polymerization to achieve polymerization directly from the surface of living cells.^[45] In this work, a lipid-based CTA was developed to insert into cell membranes and could subsequently be polymerized by utilizing EY as the catalyst to graft poly(ethyleneglycol) acrylate chains with visible light irradiation in a matter of minutes.^[45] These examples, among many others, depict the significant opportunity aqueous RAFT polymerization can provide to integrate the polymerization process with biology to achieve unique architectures and functions.

Figure 7.

a) Synthetic strategy toward aqueous RAFT of single stranded DNA functionalized polymers and nano-objects. b–e) Atomic force microscopy (AFM) images recorded by liquid AFM after aqueous RAFT dispersion polymerization from CTA modified DNA using a [DAAm]/[DMA] ratio of 80:20. Different degrees of polymerization were targeted: b) DP = 50, c) 100, d) 200, e) 250. The magnified images in b) and c) are 2.5 times magnified with respect to the original image. Reproduced with permission.^[43] Copyright 2020, Wiley-VCH.

Because of its gentle reaction conditions, aqueous RAFT has attracted much attention for its use in bioapplications via the formation of nanoparticles. Nanoparticles can be utilized in biomedicine for drug delivery, bioimaging, and diagnostic applications.^[46] Because aqueous RAFT PISA can form nanoparticles with different morphologies at high solids content, it has been utilized to prepare potential biomedically-useful materials. For example, He et al. prepared cross-linked NIPAM/allyl acrylamide vesicles via aqueous RAFT PISA loaded with the enzyme horseradish peroxidase.^[47] The membrane permeability could be regulated by changing the solution temperature, allowing for a tunable release of the enzyme cargo. By utilizing a disulfide cross-linker, dual thermo-/redox-responsive vesicles were prepared, adding additional tunability to the system. In 2020, Sun et al. reported the one-pot synthesis of spherical micelles via PISA consisting of proapoptotic peptide-polymer amphiphiles (Figure 8).^[48] In this work, it was demonstrated that the peptide-nanoparticle composites afforded higher resistance of the proapoptotic peptides against protease degradation, resulting in the nanoparticles exhibiting enhanced apoptosis efficiency in relevant in vitro models.

Figure 8.

Schematic illustration of the one-pot aqueous photo-PISA approach to proapoptotic peptide brush polymer nanoparticles. Reproduced with permission.^[48] Copyright 2020, Wiley-VCH.

3.3. Surface Initiated Polymerization

Surface-Initiated PET-RAFT (SI-PET-RAFT) was first presented in 2019 utilizing organic solvents where polymer chains were grafted directly from CTAs tethered to the surface of silica, displaying all of the hallmarks of controlled radical polymerization.^[49] Due to the advantages of PET-RAFT, this was achieved under ambient conditions and in the presence of oxygen. This afforded great spatiotemporal control of the polymerization rate by simply changing the intensity of the light source, allowing for, in the presence of a photomask, exquisite surface patterning capabilities with the polymer.^[49] This work, although achieved under bulk monomer conditions, clearly demonstrated the potential utility of SI-PET-RAFT inspiring further work utilizing aqueous RAFT surface initiated polymerization in more recent years.

SI-PET-RAFT has been utilized in aqueous solution to prepare antifouling films,^[50] and superhydrophilic films for antifogging applications.^[51] Because of its oxygen tolerance, SI-PET-RAFT is amenable to high-throughput approaches. For example, Ng et al. utilized SI-PET-RAFT to prepare a library of surface-tethered homo, statistical, gradient, and block copolymer brushes on glass surfaces to evaluate their performance in antifouling applications.^[50a] Pester and co-workers utilized aqueous SI-PET-RAFT to prepare durable films for antifogging applications completely in aqueous solution under ambient conditions and yellow light irradiation (Figure 9).^[51] In this work, the utilization of water as a sole solvent overcomes certain drawbacks compared to traditional surface-initiated radical polymerization techniques like low hydrophilic monomer solubility in organic solvents and low oxygen tolerance important for the development of superhydrophilic surfaces.^[51] Additionally, because the polymer chains are tethered to the surface, they were more durable than typical superhydrophilic coatings which can often dissolve and wash away in water with time during application.^[51]

Figure 9.

Schematic of aqueous SI-PET-RAFT for the preparation of antifogging polymer brush films demonstrated here with superhydrophilic poly[2-(methacryloyloxy)ethyl]trimethylammonium chloride (pMETAC) polymer brushes, with fogging observed for the plain glass slide (left) and maintained transparency of the pMETAC polymer brush-modified slide (right). Reproduced with permission.^[51] Copyright 2021, American Chemical Society.

3.4. High Throughput and Scale-Up Opportunities

Arguably, one of the main advantages that makes aqueous RAFT appealing for potential industrial applications is the cost, ease of access, and environmentally friendly characteristics of using water as the primary solvent. Recent reports have demonstrated the utility of aqueous RAFT to prepare libraries of polymer composition and architectures to achieve high throughput screening of polymer structures. One of the limitations of high throughput techniques for free radical polymerization has been the increased susceptibility toward dissolved oxygen. However, alternate degassing techniques such as enzyme degassing have been employed to circumvent this issue,^[52] or, as discussed previously, oxygen tolerant PET-RAFT with a suitable reducing agent (cocatalyst), making it attractive for high throughput approaches. For example, Boyer et al. prepared a series of polyacrylates and poly-acrylamides with complete monomer conversion within minutes via a continuous flow process open to the air, utilizing EY and triethanolamine as a PET-RAFT catalyst/cocatalyst system.^[53] Several diblock and triblock copolymers could be synthesized by this approach, and the molecular weights could be modified through manipulation of the flow rates, concentration, and light intensity.

Another technique for high throughput aqueous RAFT is the concept of “ultra-fast” RAFT polymerization, whereby a thermal initiator with a high radical flux is utilized in a thermal RAFT polymerization at an elevated temperature. In their pioneering work on the topic, Perrier et al. reported that the combined effects of the initiator consuming dissolved oxygen and the fast polymerization rate led to the rapid production of low dispersity polyacrylamide homopolymers and block copolymers directly in water with quantitative conversion.^[54] Further, they were able to apply this technique to a high throughput approach using only equipment typically found within biological laboratories utilized for polymerase chain reactions to establish a protocol for preparing libraries of biologically-useful acrylamide-based homopolymers, block, and statistical copolymers.^[55] Finally, in pioneering work by Sumerlin and co-workers, they demonstrated catalyst-free, aqueous RAFT photopolymerization conditions that facilitated the synthesis of ultrahigh molecular weight polymers utilizing only a readily available and low-energy light source or, in some cases, sunlight.^[56] Control of ultrahigh molecular weight polymers (defined as molecular weights greater than 1.00 ×10⁶ g mol⁻¹) with aqueous RAFT methods provides the opportunity for material properties of unrivaled mechanical strength while maintaining control of composition and architecture, highly advantageous to industrial applications of RAFT synthesized polymeric materials.

4. Conclusions and Outlook for Aqueous RAFT Polymerization

Aqueous RAFT is a powerful synthetic tool that takes advantage of the increased control of the polymerization and architectures provided by RAFT, coupled with the aforementioned advantages of water as a solvent. Although challenges arise from working in the aqueous medium, there have been many advances to circumvent these challenges as described herein. Most recently, aqueous RAFT methods have seen advances in applications for ultrahigh molecular weight products, self-assembly from stimuli-responsive polymers, and grafting polymer chains from biomolecules while maintaining their bioactivity. Further developments have also been made in the presence of oxygen utilizing continuous flow photopolymerizable systems improving the current state of the art.^[53,57] These systems use flow reactors to increase the rate of polymerization and overall yield without the loss of control while maintaining livingness.

A key highlight of aqueous RAFT is the safety and accessibility of the technique allowing for controlled radical polymerization methods highly suitable to the classroom at all levels when you consider the technique is achievable with commercially available, relatively cheap reagents, sunlight, and water as the solvent.^{[22c, 56]} Challenges remain in monomer and CTA solubility in aqueous solvents, which limits the systems that can be developed. Beyond solubility, side reactions of aminolysis and hydrolysis significantly reduce the available CTAs that can be applied in aqueous RAFT. Further investigations into optimized CTAs for performance in aqueous environments will advance applications and performance of aqueous RAFT methods.

We see tremendous opportunity in the application of aqueous RAFT in bio-relevant settings and anticipate major growth in this area. Recent studies have demonstrated a significant immune response to polyethylene glycol (PEG), a biocompatible polymer routinely used in protein and nanoparticle-based therapeutics, which can ultimately lead to adverse reactions and reduced therapeutic efficacy.^[58] As a result, biocompatible polymers of novel composition and architecture, which can potentially replace PEG in current nanomedicines and therapeutics is an important problem for the community to address, of which aqueous RAFT is perfectly poised to achieve. Further, several novel CTAs have been developed for biological applications allowing for facile bioconjugation of polymers to biological molecules and surfaces (e.g., biotinylated, lipid-based interactions, or click chemistry). The ability to synthesize polymers by aqueous RAFT and then modify the end functionality of polymer chains for in situ interactions in the biological setting we feel is an exciting prospect to be explored, facilitated by biocompatible aqueous RAFT conditions.

A major challenge currently facing society is that of plastic waste in our natural environment. RAFT mediated depolymerization, such as the recent work from the Anastasaki group to be achieved in almost quantitative yields without the need for an external catalyst,^[59] can potentially play a significant role in tackling this problem. Further work from the Haddleton group has demonstrated the ability for effective in situ depolymerization under aqueous conditions through a copper catalyzed process.^[60] We anticipate aqueous RAFT depolymerization will also play a vital role in tackling this plastic waste problem in the future. Especially when considering these depolymerization methods often requires low polymer concentrations, making water an ideal solvent as it is a safe, cheap, and environmentally friendly alternative to comparable organic solvents. Finally, given that aqueous RAFT technology is still evolving with potential applications expanding on bioconjugation, biocompatible polymerizations, stimuli-responsive copolymers, and depolymerization among others, we are excited to see an increase in industrial partnerships to support this work translating beyond the academic setting to provide a greater impact and benefit for society.

Biographies

Alexander W. Fortenberry earned a B.S. and M.S. in Chemical Engineering from the University of Mississippi under the supervision of Prof. Adam Smith. He is currently a graduate student at the University of Mississippi pursuing his Ph.D. on stimuli responsive block copolymers under the mentorship of Prof. Adam Smith.

Penelope E. Jankoski earned a B.S.E. in Polymer Science and Engineering from Case Western Reserve University in 2021. She is currently pursuing a Ph.D. in Polymer Science and Engineering from the University of Southern Mississippi under the mentorship of Prof. Tristan Clemons working on applications of polymers for tissue regeneration.

Evan K. Stacy received his bachelor’s degree from Hanover College where he studied Chemistry and Biology. Evan is now pursuing a Ph.D. at the University of Southern Mississippi’s School of Polymer Science and Engineering where he works under the advisement of Prof. Tristan Clemons working on polymeric drug delivery applications.

Charles L. McCormick received a B.S. in Chemistry from Millsaps College and a Ph.D. in Organic Chemistry from the University of Florida under the direction of George Butler. He is currently Professor Emeritus at the University of Southern Mississippi, having served 45 years with a joint appointment in Polymer Science and in Chemistry and Biochemistry. During this time, Charles graduated 56 Ph.D. students and published over 300 manuscripts in the areas of synthetic polymer chemistry, controlled free radical polymerization, responsive water-soluble polymers, and of course aqueous RAFT.

Adam E. Smith earned a B.S. and M.S. in Chemical Engineering from the University of Mississippi and a Ph.D. in Polymer Science and Engineering from the University of Southern Mississippi under the direction of Prof. Charles McCormick. He is currently Chair of Chemical Engineering and Academic Director of General Engineering at the University of Mississippi with research interests in utilizing RAFT polymerization to synthesize stimuli-responsive block copolymers for drug and gene delivery applications.

Tristan D. Clemons earned a B.S. with Honors in Nanotechnology from Curtin University and a Ph.D. in Chemistry from the University of Western Australia under the supervision of Prof. Iyer Swaminathan. Following his Ph.D. he was awarded a research fellowship from the National Health and Medical Research Council of Australia, and then relocated to Chicago to join Prof. Samuel Stupp at Northwestern University as an American Australian Association post-doctoral scholar. He is currently an Assistant Professor of Polymer Science and Engineering at the University of Southern Mississippi, with research interests in aqueous RAFT polymerization, and polymeric materials for biomedical applications.