Phương pháp atom transfer radical polymerization dịch là gì năm 2024

The development of photoinduced organocatalyzed atom transfer radical polymerization (O-ATRP) has received considerable attention since its introduction in 2014. Expanding on many of the advantages of traditional ATRP, O-ATRP allows well-defined polymers to be produced under mild reaction conditions using organic photoredox catalysts. As a result, O-ATRP has opened access to a range of sensitive applications where the use of a metal catalyst could be of concern, such as electronics, certain biological applications, and the polymerization of coordinating monomers. However, key limitations of this method remain and necessitate further investigation to continue the development of this field. As such, this review details the achievements made to-date as well as future research directions that will continue to expand the capabilities and application landscape of O-ATRP.

Graphical Abstract


1. INTRODUCTION

Atom transfer radical polymerization (ATRP) is a powerful controlled radical polymerization (CRP) method for the synthesis of polymers with targeted molecular weights, narrow molecular weight distributions (low dispersity [Đ]), varied chemical compositions, and complex architectures. In this method, a catalyst mediates the reversible activation and deactivation of polymers with halide end-groups, wherein the halide is removed during activation to generate a reactive polymer radical and then reinstalled during deactivation to yield a “dormant” polymer chain. As a result of this reversible deactivation mechanism, bimolecular radical termination reactions are suppressed, and polymer growth is controlled toward the synthesis of well-defined macromolecules with a range of functionalities and architectures.-

For the purposes of this review, it is beneficial to define certain metrics of polymerization control that are often considered in ATRP. Commonly, properties such as molecular weight control, Đ, and initiator efficiency (I*) are analyzed to evaluate control over a given polymerization. Molecular weight control refers to the ability of the user to target and synthesize polymers of varied molecular weights, commonly through manipulation of the reaction stoichiometry.

For Đ, the desirable range in ATRP and other CRPs is between 1.0 and 1.5. The lower limit of Đ = 1.0 represents a totally uniform molecular weight distribution (i.e., a single molecular weight is present). Because this value generally cannot be obtained using synthetic chemistry, values as close as possible to 1.0 are sought to indicate a well-controlled polymerization process. Instead, the upper limit of Đ = 1.5 represents the lowest Đ theoretically obtainable through a free radical polymerization process. As such, a polymerization that produces a Đ > 1.5 is not considered controlled because a similar Đ could be obtained through an uncontrolled free radical polymerization. In this review, 1.0 < Đ ≤ 1.1 is considered excellent, 1.1 < Đ ≤ 1.3 is considered good, 1.3 < Đ ≤ 1.5 is considered moderate, and Đ > 1.5 is considered poor.

Finally, I* represents the theoretical number-average molecular weight ([Mn,theo], based on the reaction stoichiometry) divided by the experimentally determined number-average molecular weight (Mn,exp). When I* = 100%, this value indicates that all the initiators added to the polymerization reaction initiated a single polymer chain, providing the user with control over the molecular weight of the polymer product. However, if I* ≠ 100%, this value can indicate that side reactions or other undesirable processes are present that may reduce control over the product polymer structure.

In seminal ATRP reports by Matyjaszewski and Sawamoto, Cu and Ru catalysts, respectively, were chosen to mediate the polymerization process. In each case, polymers could be obtained with predictable molecular weights and Đ < 1.5, indicating the radical polymerization process had been controlled to some degree. In the following years, notable advancements included the development of methods to dramatically lower catalyst loadings in ATRP,, as well as strategies to control the polymerization process using external stimuli such as light or electricity.- In particular, work reported by Fors and Hawker showed a common Ir photoredox catalyst (PC) could be used to mediate the polymerization of methyl methacrylate (MMA), providing the first example of a photoredox catalyzed ATRP method. However, around the same time, concerns surrounding the sustainability of Ir and Ru compounds began motivating the use of organic molecules as more sustainable alternatives to these catalysts.- As such, shortly after the report by Fors and Hawker, the first examples of organocatalyzed atom transfer radical polymerization (O-ATRP) emerged, employing organic PCs to mediate the polymerization of methacrylate monomers via an ATRP mechanism.,

Because of challenges associated with the reduction of C–X (X = halide) bonds (i.e., the polymer chain-end groups in ATRP), early catalyst systems for O-ATRP primarily focused on PCs that could operate by oxidative quenching of the excited state (PC*). In other words, these early catalysts systems featured strongly reducing excited states that could directly reduce the alkyl halide (Figure 1a), leading to the formation of a polymer radical (Pn•) and the catalyst radical cation (PC•+). It is hypothesized that deactivation of the polymer radical is mediated by PC•+, which in turn regenerates the neutral ground state of the catalyst (PC). Thus, the general mechanism of ATRP is maintained, but it is mediated using a photoredox catalytic cycle.


Overview of O-ATRP: (a) the mechanism of O-ATRP by oxidative quenching and common PCs employed therein, (b) the mechanism of O-ATRP by reductive quenching and a common PC family employed in this method, (c) monomer families that have been polymerized by O-ATRP, and (d) common applications of O-ATRP.

In the early development of O-ATRP, much attention was given to the development of strongly reducing PCs, such as phenothiazines, dihydrophenazines, and phenoxazines (Figure 1a). However, work by Park and Choi showed that similar polymerizations could also be performed using reductive quenching PCs such as Ru(bpy)32+ (bpy = 2,2′-bipyridine). Zhang and Cheng quickly showed this method could also be mediated by organic reductive quenching PCs (Figure 1b), in which PC* is reduced by an electron donor (D) to generate a catalyst radical anion (PC•−) and the donor cation (D+). Because PC* in these cases is often incapable of directly reducing the alkyl halide, PC•− is formed through the reaction of PC* with a sacrificial electron donor. In turn, activation mediated by PC•− generates the neutral PC, and it is proposed that the D+ mediates deactivation. As a result, common organic PCs such as xanthenes (Figure 1b) can also be used to mediate O-ATRP. While there are certain advantages and disadvantages associated with each class of PCs, these considerations will be discussed in greater detail later in the text (see , ).

Relative to other polymerization methods, O-ATRP features several desirable properties that have contributed to its popularity over time. For example, like other ATRP methods,, O-ATRP features a simple reaction setup, can produce well-defined polymers, and is tolerant to a wide range of functional groups (Figure 1c). Thanks to the use of photoredox catalysis to drive this method, O-ATRP also enjoys added benefits such as mild reaction conditions (e.g., performed at ambient temperatures) as well as spatial and temporal polymerization control through manipulation of the light source in the reaction. In addition, O-ATRP has been employed to access numerous interesting applications, including the synthesis of polymers and copolymers with complex architectures, the functionalization of various surfaces through surface-initiated polymerizations, and the production of materials for electronic and biological applications (Figure 1d).

With that said, O-ATRP has experienced several limitations since its inception that continue to attract research efforts. Regarding the mechanism of O-ATRP, significant advancements have been made in understanding PC photophysics, how PC design impacts these photophysical processes, and how these properties can affect activation during the polymerization. However, certain aspects of the O-ATRP mechanism remain poorly understood, especially in the presence of reductive quenching PCs. In addition, while significant advancements have been made in the scope of monomers successfully polymerized by this method, several of these monomers remain poorly controlled. Thus, continued research in this area is necessary to overcome the current limitations of O-ATRP and expand the utility of this method.

To promote progress in this field, this review will provide a comprehensive overview of the development, current status, and applications of O-ATRP. We begin by placing O-ATRP in context relative to other metal-free CRPs to demonstrate why one might choose O-ATRP over other, similar methods. Next, we provide a detailed account of the development of O-ATRP, including its history, common variations of O-ATRP, catalysts reported for this method, and insights gained into its mechanism. With the history of the method in mind, the current status of the field is discussed, including monomers and applications accessible through O-ATRP. Finally, we conclude this review with an opinion overviewing future directions that could expand the capabilities and utility of this method. While other reviews have been written on this topic,- they have typically focused on specific aspects of this method (ex., catalyst design, applications, etc.) rather than providing a complete overview of its development and uses. As such, this review seeks to document O-ATRP in detail, providing a comprehensive discussion of developments and advancements in the field for both new and veteran practitioners.

2. METAL-FREE CONTROLLED RADICAL POLYMERIZATIONS: O-ATRP IN CONTEXT

While the focus of this review is on O-ATRP, several other metal-free CRPs also exist that warrant comparison. Each method has its own advantages and disadvantages, so this section seeks to place O-ATRP in context relative to these other methods. Because this review is not intended to provide a comprehensive overview of metal-free CRPs, this section will focus only on common methods, including photoinduced electron/energy transfer reversible addition–fragmentation chain transfer (PET-RAFT), photoiniferter polymerization, nitroxide mediated polymerization, and iodine transfer polymerization.

Beginning with PET-RAFT, this method is the photo-catalyzed variant of the more traditional RAFT polymerization first reported by Moad, Rizzardo, and Thang in 1998. As such, PET-RAFT operates by a reversible chain transfer mechanism, where polymer growth is controlled by transferring the propagating radical from one polymer chain to another (Figure 2). In 2014, Boyer and co-workers showed the RAFT process could be mediated by common photoredox catalysts, such as fac-[Ir(ppy3)] (ppy = 2-phenylpyridine) and Ru(bpy)3Cl2. While these methods were not metal-free, it was quickly shown that PET-RAFT could also be mediated by organic catalysts, such as 5,10,15,20-tetraphenylporphyrin and related derivatives.


Mechanism of PET-RAFT proceeding through electron transfer.

Following these seminal reports, the scope of catalysts for PET-RAFT has blossomed and now includes many common organic PCs such as eosin Y, fluorescein, and more. This feature is one of the great advantages of PET-RAFT, as it can be performed with a wide variety of catalysts, many of which are commercially available, under irradiation spanning the entire visible spectrum and even into the near-infrared spectrum, which is less common in O-ATRP. Comparatively, the number of catalysts available for O-ATRP is smaller due to the greater thermodynamic requirements for activating carbon halide bonds, making catalyst selection an advantage in PET-RAFT. In addition, a large number of RAFT chain transfer agents (CTAs) have been reported over the years, and many are now commercially available, lowering the barrier to use for this method. With that said, CTAs for RAFT are generally more expensive than the alkyl halide initiators commonly employed in O-ATRP, although this can sometimes be offset by the cost of the catalyst (Table 1). In addition, because many CTAs are capable of absorbing visible light, their presence on the chain-ends of the product polymer can impart color to the polymer that may be undesirable. As a result, extra steps may be necessary to alter the polymer chain-ends and remove this color from the product.

Table 1.

Comparative Characteristics of Common Metal-free Controlled Radical Polymerization Techniques

methodreagents requiredcommercial availability of reagentscost of reagents (g−1)amonomer scopeO-ATRPphotocatalyst (PC)some PCs available, most must be synthesized$500–1000acrylamides, acrylates, acrylonitrile, methacrylates, styrene, vinyl cyclopropanes, 4-vinylpyridinealkyl halide initiatornumerous alkyl bromides and chlorides available for purchase$1–15PET-RAFTPCnumerous PCs available$10–1200acrylamides, acrylates, methacrylates, styrene, vinyl acetateRAFT agentnumerous dithiobenzoate, dithiocarbonate, and trithiocarbamate RAFT agents available for purchase$100–300photoiniferterIniferter (light absorbing RAFT agent)numerous dithiobenzoate, dithiocarbonate, and trithiocarbamate RAFT agents available for purchase$100–300acrylates, methacrylates, styrene, vinyl acetateNMPnitroxide or alkoxyaminesome nitroxides and alkoxyamines available, most must be synthesized$30–600acrylamides, acrylates, cyclic ketene acetals, 1,3-dienes, styrenes, methyl methacrylate, vinyl acetate, vinyl chlorideITPinitiatornumerous radical initiators available$1–40acrylates, methacrylates, styrene, vinyl acetate, vinyl chloridealkyl iodide or alkyl bromide + NaIlimited availability of alkyl iodides, most must be synthesized$1–100bnumerous alkyl bromides available$1–15

Closely related to PET-RAFT, iniferter polymerizations can also use CTAs to control polymer growth but without an added photoredox catalyst. Instead, the CTA is directly activated by irradiation to generate reactive radicals, which then initiate and drive the polymerization process. As a result, iniferter polymerizations are remarkably simple because a single reagent acts as the initiator, chain transfer agent, and terminator (the ini-ferter). In turn, this simplicity can greatly reduce the need for purification of the polymer product, which is often necessary with other polymerization methods such as O-ATRP and PET-RAFT. When these polymerizations are driven by photolysis of the iniferter, they are referred to as photoiniferter polymerizations (Figure 3). Importantly, while some RAFT agents can function as iniferters, it should be noted that not all iniferters can also serve as CTAs in RAFT polymerizations. One notable example is tetraphenylethane derivatives, which can serve as iniferters but not RAFT CTAs.


Origin of the name “iniferter” (left) and mechanism of a photoiniferter polymerization using dithiocarbonyl compounds (right).

Unfortunately, the simplicity of this process can also come at a cost. For instance, the use of high-energy UV light is generally undesirable in organic synthesis because it can cause side reactions that are less likely under visible light irradiation. In 2002, such side reactions were observed with certain iniferters, which were shown to undergo decomposition under polymerization relevant conditions. To circumvent this issue, one can use visible light absorbing or thermally activated iniferters,, although these constraints may also introduce added complexities by changing the chemistry of the iniferter or the polymerization process. For this reason, it is sometimes easier to employ other techniques, such as O-ATRP or PET-RAFT.

Another popular CRP method is nitroxide mediated polymerization (NMP), which employs alkoxyamines to control polymer growth through a reversible deactivation process similar to that found in ATRP. Because NMP typically involves direct homolysis of the polymer chain-end C–O bond (Figure 4), it is often performed in the absence of a catalyst. Further, the alkoxy amine can also serve as the initiator in this method, allowing these polymerizations to be performed with minimal reagents and reducing the need for polymer purification. Most commonly, NMP is performed at elevated temperature using a thermal initiator or thermally activated alkoxyamine initiator,,- although photomediated NMP has also been reported.,


Mechanism of activation and deactivation in a nitroxide mediated polymerization.

While a number of alkoxyamines have been reported for NMP,, most are not commercially available and must be synthesized prior to the polymerization. In addition, many of these compounds are unstable at elevated temperatures. As a result, they must be carefully stored and can sometimes decompose under polymerization conditions, introducing added complications relative to O-ATRP. Nonetheless, NMP remains a powerful controlled polymerization method, especially for sensitive applications where polymer impurities can be detrimental.

Finally, one CRP that is receiving increasing levels of attention is iodine transfer polymerization (ITP). While a number of ITP methods have been reported and discussed elsewhere, here we will focus on two ITP mechanisms operating by degenerative chain transfer (Figure 5a) and reversible complexation (Figure 5b). In the first, which was first reported by Tatemoto and later expanded by Matyjaszewski and co-workers in 1995, iodine is transferred from a deactivated polymer to a propagating polymer, such that polymer growth is controlled by distribution of the propagating radical across several polymer chains. As a result, the reaction system is quite simple, as the only reagents necessary for the polymerization to proceed are the monomer, initiator, and an alkyl iodide that also ultimately serves as an initiator. However, the chain transfer rate of the alkyl iodide must be matched to the propagation rate of the monomer, which can complicate reaction design.


Two common mechanisms of iodine transfer polymerization without (a) and with (b) amines as complexing agents.

Instead, ITP mediated by reversible complexation employs a complexing agent, often an amine, to assist in the removal of iodine from the polymer chain end to enable propagation. As a result, this mechanism resembles one of reversible deactivation, where the polymer is activated by removal of the iodine and deactivated by regeneration of the C–I bond. When the exchange frequency of the iodine is the limiting factor in a polymerization, this strategy can improve polymerization outcomes by facilitating iodine exchange. However, this method also introduces new reagents to the polymerization, which can complicate the reaction setup and necessitate further purification of the product polymer.

One limitation of ITP is the lack of commercially available alkyl iodides, possibly because these compounds are often thermally and photochemically unstable. One way this issue has been addressed is through in situ generation of the alkyl iodide, where alkyl bromides are converted to alkyl iodides through reaction with an iodide source (ex., NaI). Because numerous alkyl bromides are available commercially, this approach can circumvent issues related to the availability of alkyl iodides, although it also complicates the polymerization process. In addition, one must consider the stability of the polymer chain-end, which can also be susceptible to degradation either during or after the polymerization.

In summary, each polymerization method presented here features different advantages and disadvantages that one must consider in choosing a method. Perhaps one final feature that should be considered is the monomer scope of each method (Table 1), which may eliminate certain methods depending on the materials one wishes to produce. However, regardless of the specific method chosen, a central theme emerges: metal-free CRPs enable facile access to a range of polymeric materials with tunable compositions, structures, and functionalities. For instances in which O-ATRP may be the best choice, the following sections will provide deeper insight into the mechanism, scope, and applications of this method.

3. MECHANISMS OF O-ATRP

3.1. General Mechanism of ATRP

Regardless of the identity of the catalyst, all ATRP methods operate by the same general mechanism of reversible activation and deactivation (Figure 6). During activation, the catalyst (Catn) reduces an initiator molecule or a polymer chain-end possessing a C–X bond (X = halide) to generate a reactive, carbon-centered radical. Because this radical can react with functionalities within the monomer, such as alkenes, polymerization propagation can occur to grow the polymer chain. However, radical polymerizations are inherently susceptible to termination processes, wherein the propagating radicals undergo irreversible side reactions such as radical chain-coupling or disproportionation. From the standpoint of precision polymer synthesis, these side reactions are undesirable because they reduce the user’s ability to control the structure of the polymer product.


General mechanism of ATRP and key mechanistic steps. For simplicity, only one possible termination pathway is depicted (termination by combination).

To overcome this limitation, a key feature of ATRP is the formation of a deactivator during the activation step. Often, the deactivator is simply the oxidized catalyst (Catn+1), although some examples will be presented later where this may not be the case (see , ). Regardless of the identity of the deactivator, this species mediates the deactivation step of ATRP, in which the C–X polymer chain-end group is reinstalled to generate a “dormant” polymer chain and lower the concentration of radicals in solution. As a result, both the rates of propagation () and termination () are lowered. However, because radical termination reactions are typically bimolecular, the rate of termination is reduced to a greater degree than propagation. Therefore, the net result of deactivation is that termination reactions are suppressed while allowing polymer growth to proceed as desired. Importantly, for effective deactivation to occur, this step should generally be faster than the other steps of the polymerization (i.e., Rdeact > Ract, Rprop, Rterm). For further information on the general mechanism of ATRP, we point the reader to other reviews already written on this topic.,,

Rprop=kprop[Monomer][Pn⋅]

(1)

3.2. Oxidative and Reductive Quenching Mechanisms

In O-ATRP, organic PCs are used to mediate the ATRP mechanism. As such, it is important to consider the role of the PCs in addition to that of the polymerization. Most commonly, PCs that operate by oxidative quenching (Figure 7, bottom) are employed because such PCs possess strongly reducing excites states [Ered(PC•+/PC*) ≤ −1.5 V vs saturated calomel electrode (SCE)] that are capable of directly reducing the alkyl bromide or chloride in O-ATRP [E°(C–Br/C–BH•−) ~ −0.8–0.6 V vs SCE, and E°(C–Cl/C–Cl•−) ~ 0.1–0.2 V < E°(C–Br/C–BH•−)]. This mechanism begins when the PC becomes photoexcited by absorption of light to generate the strongly reducing excited state (PC*). This excited state can react with an alkyl halide, either the initiator or the polymer chain-end, to generate the reactive radical for propagation. In addition, the PC radical cation (PC•+) is formed along with a halide anion (i.e., Br− or Cl−). Together, these ions are used to deactivate the propagating radical, effectively lowering the concentration of radicals in solution and thereby limiting termination reactions. Common PC families that typically operate by this mechanism include phenothiazines, dihydrophenazines, phenoxazines, and dihydroacridines.,


Mechanisms of O-ATRP proceeding through oxidative and reductive quenching mechanisms.

PCs that operate by reductive quenching (Figure 7, top) are typically insufficiently reducing in PC* to directly reduce the alkyl halide. To circumvent this issue, sacrificial electron donors, such as amines, are oxidized by PC*, generating a more reducing PC radical anion (PC•−). If PC•− is thermodynamically capable of reducing the alkyl halide, activation can proceed. However, because the product of activation is now the neutral PC ground state, deactivation must be mediated by another species. Often, the radical cation of the sacrificial electron donor may be sufficient.

By comparing these two mechanisms, some advantages and disadvantages with each one become apparent. With oxidative quenching, fewer reagents are required. As a result, the likelihood of side reactions occurring is lowered, and contamination of the polymer product is minimized. However, PCs with strongly reducing excited states are necessary to mediate this mechanism, and such PCs, in particular strongly reducing organic PCs, have historically been rare.,,, Further, while catalyst development in O-ATRP has greatly expanded, the availability of strongly reducing organic PCs, only a handful are commercially available.-

By contrast, many of the PCs commonly employed for reductive quenching, often xanthenes, are commercially available, greatly reducing the barrier to performing O-ATRP by this photoredox mechanism. However, the requirement for a sacrificial electron donor increases the complexity of the reaction and creates new opportunities for side reactions to occur, such as undesirable hydrogen atom abstractions when amines are used, potentially limiting control over the product polymer structure. Such side reactions may be the reason that better polymerization control is often seen in O-ATRP using oxidative rather than reductive quenching PCs, although similar levels of control are possible in some cases (see , ). One final consideration is that a sacrificial electron donor may remain in the polymer as an impurity after the polymerization is complete, potentially increasing the need for polymer purification depending on the desired application.

3.3. Photocatalysts Employed in O-ATRP

3.3.1. Initial Photocatalysts.

The first examples of O-ATRP were reported simultaneously by the Miyake and Theriot and the Hawker group in 2014. In the former, perylene (Figure 8) was used as a PC and could activate ethyl α-bromophenylacetate (EBP) and catalyze the polymerization of MMA to produce poly(methyl methacrylate) (PMMA) under visible light irradiation. While this catalyst gave only moderate polymerization control, Đ as low as 1.3, initiator efficiency (I*) ∆ 100%, it also provided the first example of performing O-ATRP using visible light. This feature is desirable because UV light is more likely to cause side reactions by direct activation of other organic molecules in solution. Further, by employing perylene as the PC, high molecular weight polymers (Mn > 100 kDa) could be produced by O-ATRP (Mn = 125–273 kDa), and several monomers (methacrylates, acrylates, and styrene) were successfully polymerized.


Structures of common PCs and PC families employed in O-ATRP. For phenothiazines, dihydrophenazines, and phenoxazines, the following naming system is used in this review: PhenX-N_aryl-core, where PhenX refers to the identity of the PC core (ex., PhenS = phenothiazine, PhenN = dihydrophenazine, and PhenO = phenoxazine); N_aryl is an abbreviation referring to the N-aryl substituent (ex. 1N = 1-naphthyl), and core is an abbreviation referring to the core substituents (ex. BiPh = 4-biphenyl). For dihydroacridine PCs, PhenX is replaced with Acrid.

To demonstrate that these polymerizations were driven by light, on/off experiments were performed in which the polymerization was repeatedly stopped and restarted by manipulation of the light source. During each “off” period, conversion remained unchanged, increasing only when the lights were turned on again. Further, a control experiment was performed using orange instead of white light, which could not be absorbed by the catalyst. Because no polymerization was observed, this experiment suggested that the reaction was in fact driven by light and not simply the heat produced by the light source.

In addition to investigating the role of light in these polymerizations, experiments were also performed to demonstrate that these reactions indeed proceeded through an ATRP mechanism. For example, kinetic experiments showed the polymerization exhibited linear pseudo-first-order kinetics; however, Đ increased throughout the polymerization while Mw decreased, the opposite of what is expected for an ATRP process. Nevertheless, investigation of the chain-end groups by matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS) showed the presence of bromine chain-ends, which enabled the polymers to be chain-extended with various monomers to produce block copolymers. As such, it was concluded that these polymerizations did in fact proceed through an ATRP mechanism mediated by perylene.

Around the same time, Hawker and co-workers reported the first use of phenothiazines to mediate O-ATRP under UV irradiation. Specifically, 10-phenylphenothiazine (PhenS-Ph, also refered to as PTH) exhibited excellent performance, producing PMMA with Mn = 1.3 to 15.4 kDa, Đ as low as 1.2, and I* ~ 100%. Further, dimethylaminoethyl methacrylate (DMAEMA), a monomer that can poison metal-based catalysts by coordination, was successfully polymerized with excellent control (Mn = 8.8 kDa, Đ = 1.1), making this the first example to highlight the utility of organocatalysis in ATRP.

By cycling the light source “on” and “off,” it was shown that this polymerization reaction was driven by light and that the reaction could be started and stopped repeatedly without loss of polymerization control. Investigation of chain-end group fidelity in the polymer products was performed using electrospray ionization mass spectrometry (ESI-MS), 1H nuclear magnetic resonance (NMR) spectroscopy, and chain-extension experiments. Together, these experiments demonstrated that this method provided excellent chain-end fidelity and that the polymers produced by O-ATRP could be further functionalized by other ATRP methods such as Ir(ppy)3 catalyzed photo-controlled ATRP (photoATRP) and Cu ATRP. Importantly, the compatibility of O-ATRP with other ATRP methods enables unique copolymers to be synthesized by leveraging the complementary strengths of different catalyst systems.

In comparing these two initial catalyst systems, each had associated advantages and disadvantages that would motivate future catalyst development. For example, the superior ability of PhenS-Ph to produce polymers with low Đ and near-quantitative I* was immediately evident and can be attributed to the superior excited-state redox properties of this catalyst (Table 2). For example, the triplet excited state of PhenS-Ph is far more reducing than that of perylene [E°(PC•+/3PC*) = −1.7 V for PhenS-Ph vs −0.58 V for perylene,- both vs SCE], which should facilitate more efficient activation of the alkyl bromide [E°(C–Br/C–Br•−) 0.8 to −0.6 V vs SCE] throughout the polymerization. However, PhenS-Ph required irradiation with UV light, whereas perylene was able to operate under visible light irradiation. While UV irradiation was not detrimental in this seminal report, it is more susceptible to initiating side reactions in certain systems. To avoid this possibility, the use of visible light to drive the polymerization is more desirable. As is outlined in the next two sections, addressing this disparity through the design of visible light absorbing, strongly reducing organic PCs was a significant research focus during the early development of O-ATRP.

Table 2.

Redox Properties of Photoredox Catalysts Used in O-ATRP

ground state

excited state (S1)

excited state (T1)

photocatalyst (PC)solventE1/2 (PC+/PC)E1/2 (PC−/PC)E° (PC+/1PC*)E° (PC−/1PC*)E° (PC+/3PC*)E° (PC−/3PC*)refPhenS-PhMeCN0.68−2.1−1.7 DMAc0.82−1.97 PhenS-1NDMAc0.83−2.23 PhenS-BiPh-t-BuPhDMF0.76−1.94 b-PhenS-PhDMAc0.90−1.92 PhenN-PhCF3MeCN0.32−1.80−2.06a PhenN-1NMeCN0.23−1.64−2.12a PhenN-PhCF3-2NDMAc0.38−1.84−1.79a PhenO-1N-BiPhDMAc0.65−1.80−1.70a PhenO-2N-PhCNDMAc0.69−1.75−1.42a PhenO-2N-MeOPhDMAc0.52−1.81−1.91a PhenO-2N-Ph3NDMAc0.54−1.83−1.88 Acrid-1N-MeOPhDMF0.71−1.73−1.62b peryleneMeCN0.98−1.87−0.58 pyreneMeCN1.24−2.12 4-CzIPNMeCN1.52−2.12 fluoresceinMeCN0.87−1.17−1.551.25−1.070.77 eosin YMeCN0.76−1.08−1.581.23−1.150.83 benzophenoneMeCN2.39−1.72−0.831.5−0.611.28 ,

3.3.2. Dihydrophenazines.

One of the first PC families to possess both strongly reducing excited states and visible light absorption was the N,N-diaryl dihydrophenazines. First introduced in 2016 by Theriot et al., the dihydrophenazine family was identified as a class of PCs through computational methods and predicted to feature excited-state reduction potentials [E°(PC•+/3PC*)] as low as −2.36 V vs SCE. As this reduction potential is more than sufficient to reduce the alkyl halides in O-ATRP, a series of four dihydrophenazines with varying N-aryl substituents was synthesized and investigated. Substituents were chosen featuring both electron donating groups (EDGs) and electron withdrawing groups (EWGs) to investigate how the electronics of the N-aryl substituent impacts PC properties, which led to the discovery of several important structure–property relationships. For example, the use of EDGs made 3PC* more reducing while also decreasing the oxidation potential of the PC radical cation [E°(PC•+/PC)]. Instead, EWGs had the opposite effect, increasing E°(3PC•+/3PC*) (i.e., making it more positive and 3PC* less reducing) and increasing the oxidation potential of the radical cation.

Interestingly, when these dihydrophenazines were employed in O-ATRP, all four effectively catalyzed the polymerization of MMA. However, PCs with EWGs exhibited superior polymerization control, producing polymers with lower dispersity (Đ ≤ 1.3) than PCs with EDGs. In particular, 5,10-di(4-trifluoromethylphenyl)-5,10-dihydrophenazine (PhenN-PhCF3, Figure 8) showed good control in the polymerization of MMA, producing PMMA with Mn = 6.0 to 58.6 kDa, Đ as low as 1.1, and I* ranging from 50% to 85%.

To understand how the identity of the N-aryl substituent impacts catalysis, density functional theory (DFT) was used to probe the excited states of each PC. Visualization of the computed singly occupied molecular orbitals (SOMOs) in 3PC* revealed that PCs with EDGs had two SOMOs both located on the dihydrophenazine core in the excited state. Instead, PCs with EWGs featured a lower-lying SOMO localized on the dihydrophenazine core and a higher-lying SOMO on the N-aryl substituent (Figure 9). This spatial separation of the SOMOs was interpreted as intramolecular charge transfer (ICT) in 3PC*, analogous to metal to ligand charge transfer in transition-metal PCs such as Ru(bpy)32+.,


SOMOs of dihydrophenazines with and without ICT in the triplet excited state.

Since ICT in 3PC* appeared beneficial for catalysis in O-ATRP, new PC targets were explored that also exhibited this computationally predicted property. As a result, two new dihydrophenazines bearing 1-naphthyl (PhenN-1N) and 2-naphthyl N-aryl substituents were identified and employed in O-ATRP. Excitingly, both PCs demonstrated excellent polymerization control in the synthesis of PMMA, supporting the importance of ICT as a key design principle for O-ATRP catalysts. Further, PhenN-2N later became commercially available, reducing the barrier to implementation of this catalyst in future reaction development.

Further functionalization of N,N-diaryl dihydrophenazines was later enabled by a synthetic breakthrough allowing for bromination of the dihydrophenazine core. By subjecting the tetrabrominated dihydrophenazine to Suzuki coupling conditions, several new PCs featuring aryl-core substituents were developed with enhanced light absorption, indicated by an increase in molar absorptivity, and further tunable redox properties. Interestingly, while several of these new compounds were still predicted to feature ICT in 3PC*, DFT calculations suggest the ICT state is localized on the core substituents rather than the N-aryl substituents. Despite this difference in ICT, the previous trend in polymerization performance was again observed; PCs with ICT excited states displayed better control in O-ATRP than PCs without ICT.

In addition, these core-extended dihydrophenazines exhibited several advantages over their noncore substituted counterparts. For example, core substitution led to improved control in the polymerization of MMA, as exhibited by the production of polymer with lower Đ and near-quantitative I*. However, perhaps the most notable improvement in catalysis was the ability of core-substituted phenazines to operate at significantly reduced catalyst loadings without loss of polymerization control. In the case of PhenN-PhCF3-2N (Figure 8), PMMA was successfully synthesized, with Mn = 7.8 kDa, Đ = 1.1, and I* = 106% using just 50 ppm of catalyst. By contrast, lowering the concentration of PhenN-PhCF3 from 1000 to 50 ppm resulted in a significant increase in dispersity (Đ = 1.3 at 1000 ppm, 1.66 at 50 ppm), indicating a loss of polymerization control at low catalyst loadings.

Despite these advances in catalyst design, one long-standing challenge in O-ATRP has been the development of catalyst systems or reaction conditions to polymerize a wide variety of monomers in a controlled fashion. Toward this end, Jessop and Cunningham reported a novel dihydrophenazine catalyst bearing amine-functionalities on the N-aryl substituents (Figure 10) that proved particularly effective in the polymerization of styrene by O-ATRP (Mn = 17.0 to 18.0 kDa, Đ = 1.1 to 1.2). This report represents the first example of polymerizing styrene by O-ATRP with excellent polymerization control. However, it remains unclear how the properties of this new PC differ from previous dihydrophenazines and how these properties contribute to the controlled polymerization of styrene. As such, further investigation of this catalyst is warranted.


Recyclable dihydrophenazine PC developed by Jessop and Cunningham.

In addition to the O-ATRP of styrene, catalyst recycling was demonstrated in both organic and aqueous polymerizations by protonation or deprotonation of the amine functionalities and extraction of the catalyst from the polymerization solution. For example, after the polymerization of styrene, carbonated water was added to the reaction to protonate and extract the catalyst from the organic phase, allowing for the catalyst to be reused in subsequent polymerizations. Similarly, when hydroxyethyl methacrylate (HEMA) was polymerized in water using the protonated catalyst, deprotonation of the PC and addition of toluene to the reaction enabled the catalyst to be recovered postpolymerization.

In an effort to further expand the utility of dihydrophenazine photocatalysts, McCarthy et al. investigated the ability of phenazine PCs to mediate the O-ATRP of acrylates. Compared to methacrylates, the polymerization of acrylate monomers is more challenging to control due to their faster propagation, which requires faster deactivation to minimize uncontrolled propagation and termination reactions. Furthermore, reduction of a secondary alkyl bromide for a dormant acrylate-based polymer is more challenging than reduction of a tertiary alkyl bromide for a methacrylate-based polymer. Interestingly, when 5,10-di(2-naphthyl)-5,10-dihydrophenazine (PhenN-2N) was employed, it was discovered that the choice of solvent could significantly impact the success of the polymerization. In general, less polar solvents improved polymerization control (lower Đ, higher I*) relative to more polar solvents, enabling the controlled polymerization of n-butyl acrylate (n-BuA). For a more detailed explanation of the impact of solvent polarity on PC properties and their ability to mediate a controlled polymerization, please see , .

It was also discovered that PhenN-2N could undergo core-substitution by the initiator during early polymerization times, leading to the in situ formation of a new catalytic species. Ultimately, the in situ generation of this new core-substituted PC prior to O-ATRP was exploited to produce well-defined poly(n-butyl acrylate) with a range of molecular weights (Mn = 7.7–17.5 kDa), low dispersity (Đ = 1.1–1.4), and moderate-to-excellent initiator efficiencies (I* = 183% to 93%). Further, this polymerization system was expanded to a number of other acrylate monomers and even to the synthesis of well-defined block copolymers with similar levels of control, demonstrating the versatility of this new catalyst system. In later work, it was shown that this core-substitution side reaction could be used to generate a series of new phenazine PCs and that core-substitution could be used to tune several catalytically relevant properties.,

3.3.3. Phenothiazines.

Following the seminal report by Hawker that employed PhenS-Ph as the PC for O-ATRP, a number of other phenothiazine PCs were developed by variation of the N-aryl substituent as well as through core-substitution. One of the first examples of this catalyst development came from Matyjaszewski in 2015 when phenothiazines with 4-methoxyphenyl (PhenS-MeOPh) and 1-naphthyl (PhenS-1N) N-aryl substituents were reported for the polymerization of acrylonitrile (AN) by O-ATRP. However, as PhenS-Ph ultimately provided the best control in the polymerization of this monomer, these new PC variants did not receive extensive use. Later in 2016, Matyjaszewski reported several new phenothiazines with N-aryl substituents such as 4-chlorophenyl and 2-pyridinyl, as well as the use of PhenS-MeOPh and PhenS-1N in the O-ATRP of MMA. This time, PhenS-1N exhibited better performance than PhenS-Ph, producing PMMA with lower dispersity (Đ = 1.4 vs 1.5), albeit at relatively low monomer conversions (10% and 16%, respectively).

Despite these efforts to tune PC properties through variation of the N-aryl substituent, all of the phenothiazine variants discussed to this point required UV irradiation during O-ATRP. To address this limitation, Matyjaszewski and co-workers developed phenyl benzo[b]phenothiazine (b-PhenS-Ph, Figure 8). Relative to previous phenothiazines, b-PhenS-Ph features a larger aromatic core, which red-shifts its absorption into the visible range. As a result, this PC was able to mediate the O-ATRP of MMA under 392 nm light, producing PMMA with Mn ranging from 7.3 to 21.7 kDa and Đ = 1.3 to 1.7. However, I* with this catalyst system was generally low, ranging from 18% to 69% for the polymerization of MMA. Nevertheless, the polymerization could be stopped and restarted several times by manipulating the light source, and block copolymers were successfully synthesized, indicating retention of the bromine chain-ends of the polymers produced.

Similarly, Chen and co-workers sought to develop visible light absorbing phenothiazines, this time using a core-substitution approach similar to that first employed with phenoxazine PCs in 2016. By installing 4-n-butylphenyl groups at the 3- and 7-positions of the PC core, phenothiazines with various N-aryl substituents were prepared that exhibited strong light absorption tailing into the visible region. In particular, PhenS-BiPh-n-BuPh (Figure 8) exhibited good control over the polymerization of MMA, producing polymer with Mn = 11.2 kDa, Đ = 1.2, and I* = 81%. This PC was also explored in the polymerization of other monomers, including acrylates and acrylamides, although these polymerizations were generally less controlled (Đ > 1.5) than those of methacrylates.

In addition to these more common phenothiazines, a number of other PCs within this family have been developed by modification of the N-aryl group. Indeed, functionalities such as 2-naphthyl, 4-cyanophenyl, 4-trifluoromethylphenyl, 4-triphenylamine, and 1-pyrenyl groups have all been incorporated into novel phenothiazine PCs. However, PhenS-Ph remains perhaps the most popular phenothiazine for O-ATRP, possibly due to its versatility and commercial availability.

3.3.4. Phenoxazines.

Another popular class of photocatalysts that has received considerable development is the phenoxazine family. First introduced in 2016 by Pearson et al., initial interest in this PC family stemmed from the desire to create strongly reducing, visible light absorbing catalysts for O-ATRP. Toward this end, phenoxazines with various N-aryl substituents were synthesized and investigated experimentally and computationally. Given the previous observation that phenazines with EWGs and extended conjugation on the N-aryl substituent provided the best polymerization control in O-ATRP, this work primarily investigated PCs with 4-trifluoromethylphenyl (PhenO-PhCF3), 1-naphthyl (PhenO-1N), and 2-naphthyl (PhenO-2N) groups. However, it is interesting to note that computations predicted PhenO-PhCF3 would not undergo ICT in 3PC* because both computed SOMOs were localized on the phenoxazine core. Supporting this prediction, PhenO-PhCF3 provided only moderate polymerization control in the O-ATRP of MMA (Mn = 6.5 kDa, Đ = 1.5, I* = 86%). Instead, PhenO-1N and PhenO-2N were both predicted to poses ICT excited states, which in turn afforded enhanced polymerization control (Mn = 8.8 kDa, Đ = 1.2, I* = 93% for PhenO-1N; Mn = 10.8 kDa, Đ = 1.1, I* = 77% for PhenO-2N).

Despite these results, initial phenoxazine PCs remained limited by their requirement for UV light to drive their reactivity. In an effort to retain their favorable catalytic properties but also red-shift their absorption into the visible spectrum, core-substitution of PhenO-1N was undertaken to install 4-biphenyl substituents at the 3- and 7-positions of the phenoxazine core. The resulting catalyst (PhenO-1N-BiPh, Figure 8) exhibited strong visible light absorption, enabling O-ATRP to be performed under white light irradiation. Further, the ability of this catalyst to mediate the controlled polymerization of MMA under various conditions was demonstrated, with polymers produced exhibiting Mn = 3.6–21.3 kDa, Đ = 1.1–1.3, and I* 98% to 105%. As such, while PhenO-1N-BiPh was not the first PC for O-ATRP with both visible light absorption and a strong excited-state reduction potential [E°(PC•+/3PC*) = −1.70 V vs SCE], it did represent the first catalyst system that could produce polymer with both low Đ and quantitative I* under visible light irradiation.

Expanding on this strategy of core-substitution, future development of the phenoxazine family resulted in the installation of a variety of core substituents with both EWGs and EDGs to tune the redox properties of the catalysts. For example, through incorporation of EWGs, E°(PC•+/3PC*) could be increased by as much as 280 mV relative to PhenO-1N-BiPh [E°(PC•+/3PC*) = −1.42 V vs SCE for 3,7-di(4-cyanophenyl)-2-naphthyl-10-phenoxazine (PhenO-2N-PhCN)]. Similarly, by destabilizing the radical cation of the PC, installation of EWGs was also found to change E°(PC•+/PC) by as much as 40 mV relative to PhenO-1N-BiPh [E°(PC•+/PC) = 0.69 V vs SCE for PhenO-2N-PhCN and 0.65 V vs SCE for PhenO-1N-BiPh]. Instead, when EDGs were installed on the phenoxazine core, a decrease in both E°(PC•+/3PC*) and E°(PC•+/PC) was observed. For example, installing 4-methoxyphenyl substituents (PhenO-2N-MeOPh) yielded E°(PC•+/3PC*) = −1.91 V vs SCE and E°(PC•+/PC) = 0.52 V vs SCE, whereas 4-triphenylamine functionalities gave E°(PC•+/3PC*) = −1.88 V vs SCE and E°(PC•+/PC) = 0.54 V vs SCE.

In addition to investigating the impact of core substituents on PC redox properties, this work also probed how different functionalities affect light absorption and ICT within the PC. For example, by systematically increasing the amount of conjugation on the phenoxazine core, it was shown that the absorption of the catalyst could be red-shifted and the molar absorptivity increased to improve visible light absorption. Moreover, computational modeling was employed to visualize the electrostatic potential (ESP)-mapped electron density in phenoxazines with various N-aryl and core substituents in an effort to understand ICT in each of these compounds. Notably, these computational results suggested that PCs with biphenyl core substituents could exhibit ICT to the biphenyl group even if the N-aryl substituent cannot support ICT. This prediction was then supported through measurement of the PC emission spectra. In the case of PhenO-2N, a broad, featureless emission was observed with a Stokes shift of 191 nm in DMAc, consistent with ICT to the 2-naphthyl group. Instead, while PhenO-2N-BiPh also displayed a broad, featureless emission, consistent with an ICT state, its Stokes shift was much smaller (82 nm). This difference in emission suggests the nature of the ICT state observed in PhenO-2N-BiPh is distinct from that present in PhenO-2N. Most interestingly, when the N-aryl group was changed from 2-naphthyl to phenyl (PhenO-Ph-BiPh), an emission spectrum nearly identical to that of PhenO-2N-BiPh was obtained, suggesting an ICT state localized on the biphenyl moiety of these PCs. Practically, the effect of this ICT state was observed in O-ATRP when PhenO-Ph-BiPh was able to produce well-defined PMMA (Đ = 1.2, I* = 102%) while PhenO-Ph could not (Đ = 1.5, I* = 111%).

While Pearson et al. and McCarthy et al. largely focused on phenoxazines with phenyl and naphthyl N-aryl groups,, Lee and Son developed a series of phenoxazine PCs with 4-trifluoromethylphenyl N-aryl substituents. Again, core-substitution with aromatic functionalities was found to red-shift the absorption of the PCs, as well as increase their molar absorptivity. Further, the installation of EWGs and EDGs was found to tune the redox properties of the PCs, consistent with the trends previously reported. In the O-ATRP of MMA, the catalyst with 4-biphenyl core substituents (PhenO-PhCF3-BiPh) was ultimately the most successful, producing PMMA with Mn = 7.5–12.9 kDa, Đ = 1.2–1.5, and I* = 49% to 92%. In addition, this PC was employed in the synthesis of amphiphilic block copolymers, where bromine-functionalized poly(ethylene glycol) (PEG-Br) was chain-extended with glycidyl methacrylate by O-ATRP.

In an effort to develop water-soluble O-ATRP catalysts, Zhou and Luo employed the core-substitution strategy to add PEG functionalities to PhenO-1N (PhenO-1N-PEG, Figure 11). Because PhenO-1N exhibits ICT to the 1-naphthyl group, this strategy should preserve the favorable catalytic properties of PhenO-1N while increasing its solubility in aqueous solvent systems. Indeed, when PhenO-1N-PEG was employed in the O-ATRP of various monomers, good polymerization control was observed. For PEG methacrylate, polymer with Mn ranging from 26.7 kDa to 67.1 kDa was obtained, with Đ = 1.2–1.3 and I* = 60% to 79%. Further, the polymer was successfully chain-extended, suggesting retention of the bromine chain-end functionality. For the analogous PEG acrylate monomer, similar results were obtained, albeit with slightly reduced polymerization control (Mn = 81.9 kDa, Đ = 1.4, I* = 51%). In addition, the polymerization of N-isopropylacrylamide (NiPAM) was also attempted, although this polymerization did not exhibit the features of a controlled radical polymerization (Mn = 71.4 kDa, Đ = 2.6, I* = 15%).


PEG-Functionalized, water-soluble phenoxazine.

By further tuning the N-aryl substituent of phenoxazine PCs, Chen and Fang developed PCs for O-ATRP exhibiting thermally activated delayed fluorescence (TADF). In essence, these PCs were developed to feature a small energy gap between S1 and T1, allowing reverse intersystem crossing (T1 to S1) to take place through thermal excitation. The result should be an increased yield of S1 that can be used to mediate activation during O-ATRP. In the polymerization of MMA, these PCs with TADF showed moderate polymerization control (Mn = 1.4–17.8 kDa, Đ = 1.3–1.9, I* = 6% to 180%).

Finally, the versatility of phenoxazines was demonstrated in 2018 when PhenO-1N-BiPh was shown to mediate the controlled polymerization of MMA under air. Typically, O-ATRP has been performed under an inert atmosphere to avoid possible interference from oxygen. For example, oxygen is known to quench propagating radicals, as well as engage in energy transfer with PCs operating from triplet excited states, because the ground state of oxygen is also a triplet. However, this work showed that under the appropriate conditions, PhenO-1N-BiPh could successfully perform O-ATRP in the presence of oxygen, greatly reducing the complexity of this polymerization system. Key to overcoming this challenge was optimization of the reaction headspace, where it was shown that eliminating the headspace of air enabled the synthesis of well-defined PMMA (Mn = 11.3 kDa, Đ = 1.2, I* = 87%), while the same polymerization with ~18 mL of headspace of air was completely uncontrolled (Mn = 7.6 kDa, Đ = 1.9, I* = 50%). Through the synthesis of block copolymers, support was found for good chain-end fidelity in these polymerizations, suggesting the presence of a small quantity of oxygen does not significantly impact the ATRP process. In addition, similar polymerization results were also obtained with several other phenoxazine PCs, suggesting this strategy can be generalized across the phenoxazine family.

3.3.5. Dihydroacridines.

One major limitation of O-ATRP has historically been its narrow monomer scope, especially in comparison to traditional ATRP methods., In an effort to expand this method’s scope, acrylates have often been targeted in O-ATRP. However, these monomers present several challenges that make their controlled polymerization difficult. In particular, the propagation rate constants for radical polymerizations of acrylates are typically an order of magnitude larger than those of methacrylates. As a consequence, faster and more efficient deactivation is necessary to control the propagation of acrylates. Additionally, the bromine chain-end groups of acrylates are more challenging to reduce,, impeding activation of the polymer. As such, overcoming these challenges requires strongly reducing PCs that can also mediate deactivation effectively.

One strategy that has been explored to achieve these properties has been the development of novel PCs similar to phenothiazines, phenoxazines, and dihydrophenazines featuring more oxidizing radical cations. The rationale behind this approach is that increasing E°(PC•+/PC) increases the driving force for deactivation, which might in turn increase the rate of deactivation in O-ATRP. Toward this end, dihydroacridines were developed in 2020, which display excited-state reduction potentials [E°(PC•+/3PC*)] ranging from −1.62 to −1.49 V and oxidation potentials [E°(PC•+/PC)] from 0.71 to 0.90 V (all vs SCE). Interestingly, while 9,9-dimethyl-10-(1-naphthyl)-9,10-dihydroacridine (Acrid-1N) does not exhibit electrochemical reversibility, indicating degradation of the radical cation, core-substitution with aryl functionalities stabilizes the radical cation. As a result, core-substituted dihydroacridines such as Acrid-1N-MeOPh (Figure 8) can undergo reversible oxidation and subsequent reduction, allowing them to operate as catalysts in O-ATRP. In addition, computational and spectro-scopic investigation of these compounds’ excited states revealed evidence of excited-state ICT, suggesting these compounds would perform favorably in O-ATRP.

To probe the catalytic properties of these new compounds, the O-ATRP of n-BuA was attempted using seven dihydroacridine PCs. It was discovered that Acrid-1N-MeOPh gave the best polymerization control, producing polymer with Mn = 10.6 kDa, Đ = 1.5, and I* = 96% under batch irradiation conditions. Because this PC was the least oxidizing [E°(PC•+/PC)] and most reducing [E°(PC•+/3PC*)] acridine investigated, it was hypothesized that a balance in redox properties was necessary to ensure efficient deactivation while maintaining the ability of the catalyst to activate the polymer chain-end.

To further improve activation in this polymerization system, it was proposed that the use of a continuous flow reactor would be beneficial because this reactor design can provide more uniform irradiation of the reaction solution than a batch reactor.- Consistent with this hypothesis, when a flow reactor was used, Acrid-1N-MeOPh produced poly(n-butyl acrylate) with nearly identical molecular weight (Mn = 11.0 kDa) and initiator efficiency (I* = 97%), but lower dispersity (Đ = 1.4). Through further optimization and the use of bromide salts, improved polymerization control was ultimately demonstrated for n-BuA (Mn = 5.4–26.4 kDa, Đ = 1.2–1.4, and I* = 44% to 173%) as well as a number of other acrylate monomers.

Through modification of the N-aryl and core substituents, Ma and co-workers developed a similar series of dihydroacridines for O-ATRP. Similar to the previous report, the electrochemical reversibility of the acridines was significantly improved by core-substitution with aryl functionalities. However, these compounds exhibited even more strongly oxidizing radical cations [E°(PC•+/PC) = 0.94 to 1.02 V vs SCE], expanding the range of redox potentials accessible by this class of organic PCs. In the O-ATRP of methacrylates, catalysts with 3′,5′-trifluoromethyl-4-biphenyl substituents exhibited the best catalytic performance, producing poly(methacrylates) with Mn = 8.9–12.2, Đ = 1.1–1.6, and I* = 53% to 105%.

3.3.6. Polyaromatic Hydrocarbons.

In addition to the phenazines, phenothiazines, phenoxazines, and acridines discussed above, a number of polyaromatic hydrocarbons (PAHs) have been investigated and employed as PCs in O-ATRP. Of course, the first example of such a catalyst in O-ATRP is perylene (see , ). Closely related to perylene are anthracene and pyrene, which were explored as organic PCs for O-ATRP by Yilmaz and Yagci in 2016. When anthracene was used under UV irradiation, the polymerization of MMA showed signs of moderate polymerization control (Mn = 4.1–19.1 kDa, Đ = 1.4–1.5). However, 1H NMR analysis of the reaction before and after irradiation suggested the presence of a side reaction with the anthracene catalyst, which the authors proposed involves substitution of the catalyst by the initiator radical followed by growth of the polymer chain from the PC core.

Instead, pyrene exhibited slightly improved polymerization control, producing PMMA with a range of molecular weights (Mn = 11.0–36.0 kDa), moderate dispersity (Đ = 1.4–2.1), and moderate initiator efficiencies (I* = 12% to 112%). Similar polymerization control was observed when pyrene was used to mediate the O-ATRP of other monomers, such as t-butyl acrylate (Mn = 107 kDa, Đ = 1.3, I* = 25%) and styrene (Mn = 2.0 kDa, Đ = 1.3, I* = 383%). Perhaps the most interesting detail about this catalyst, though, is its propensity to form excimers, or excited-state dimers. To probe the role of these excimers in catalysis, Yagci and co-workers performed Stern–Volmer quenching experiments to measure the rates of activation with various ATRP initiators. Under dilute conditions, where excimer formation is disfavored, activation rate constants ranging from 4.4 × 107 to 1.1 × 108 M−1 s−1 were obtained. Under high concentrations of pyrene, where excimer formation is favored, activation rate constants ranged from 1.6 × 107 to 7.5 × 107 M−1 s−1. As such, both species may contribute to activation during O-ATRP.

In addition to unfunctionalized PAHs such as perylene, anthracene, and pyrene, a number of functionalized PAHs have also been employed as PCs in O-ATRP. For example, 3,4,9,10-tetra-(12-alkoxycarbonyl)-perylene, an ester functionalized perylene bearing alkyl chains, was shown to mediate the polymerization of MMA under blue light irradiation. Interestingly, while PMMA with similar dispersity (Đ = 1.4–1.5) was obtained relative to when perylene was the PC (Đ = 1.3–1.9), I* was significantly improved (I* = 86% to 97%) relative to unsubstituted perylene (I* = 2% to 88%). Further, by increasing conjugation within the PAH core (Figure 12), it was demonstrated that the absorption of the PC could be red-shifted into the near-IR (NIR) spectrum. As a result, the O-ATRP of MMA could be conducted under NIR irradiation with moderate polymerization control (Mn = 2.2–8.7 kDa, Đ = 1.3–1.5, I* = 73% to 94%) and under low catalyst loadings (1–20 ppm of PC).


A NIR absorbing PC developed for O-ATRP in 2018 by Liang and Wang.

More recently, Liao and co-workers developed a strategy to synthesize heteroatom-doped PAHs through the oxidative cyclization of 1,1′-bisnaphthol (BINOL) derivatives. The resulting PAHs could be further functionalized by modification of the BINOL precursor, enabling the installation of alkyl and aryl functionalities on the PAH core. Using computational (DFT) and experimental (cyclic voltammetry) methods, the excited-state reduction potentials [E°(PC•+/3PC*)] of the new compounds synthesized were estimated to be roughly −1.6 V vs SCE, whereas the oxidation potentials of their radical cations [E°(PC•+/PC)] ranged between 0.74 and 0.82 V vs SCE. In the O-ATRP of MMA, all of the PCs synthesized provided moderate or good polymerization control (Mn = 10.9–19.9 kDa, Đ = 1.1–1.3, I* = 46% to 73%). Most excitingly, these PCs could mediate the O-ATRP of various methacrylates at concentrations as low as 0.05 ppm (Mn = 30.8 kDa, Đ = 1.5), a significantly lower catalyst loading than that employed in early O-ATRP methods (generally 1000 ppm of PC).

Yagci and co-workers also developed a series of heteroatom-containing PAHs for O-ATRP, although in this case they were based on functionalized thienothiophenes. In total, four PC derivatives were synthesized, each with increasing quantities of conjugation that served to red-shift the PC absorption into the visible spectrum. Through fluorescence quenching experiments, the authors demonstrated that one of these derivatives could undergo an excited-state reaction with ethyl α-bromoisobuty-rate, an alkyl bromide initiator for O-ATRP. However, because of the nature of these experiments, they could not inform whether this reaction proceeded through an electron transfer or energy transfer pathway. Interestingly, less functionalized derivatives exhibited better control in O-ATRP through the production of polymers with lower dispersity (Đ ~ 1.3 vs 1.7 for more functionalized PC derivatives).

3.3.7. Other Oxidative Quenching Photocatalysts.

Several other photocatalysts and dyes have also been explored as catalysts in O-ATRP. For example, Zhang and Cheng investigated 1,2,3,5-tetrakis(carbazole-9-yl)-4,6-dicyanoben-zene (4-CzIPN, Figure 8) in the O-ATRP of MMA, although this catalyst afforded only moderate polymerization control under optimized conditions (Mn = 19.1 kDa, Đ = 1.5, I* = 95%). In a later report, Kim, Gierschner, and Kwon showed 4-CzIPN could produce PMMA with Đ as low as 1.37 (Mn = 24.2 kDa), although with reduced initiator efficiency (I* = 63%).

In addition, Kim, Gierschner, Kwon, and co-workers explored a number of other PC targets using computational methods. One of these PCs, which can be described as two N-phenyl dihydrophenazines sharing a common diphenyl sulfone linker, demonstrated moderate control over the polymerization of styrene, producing poly(styrene) with Mn = 8.7 kDa, Đ = 1.4, and I* = 90%. Similarly, Zhou and co-workers developed a PC containing two phenothiazine moieties linked at the N-aryl positions by 2,7-fluorenone, which was able to produce PMMA with Mn = 3.1–10.1 kDa, Đ = 1.4–1.7, and I* = 60% to 95%.

In 2018, Wang and Zhang showed that substituted benzothiadiazoles and benzotriazoles could mediate the polymerization of MMA in the presence of a sacrificial amine. Interestingly, while the addition of an amine did improve polymerization control, the polymerization proceeded even in the absence of added amine. This observation suggests these organic PCs could activate the alkyl halide for growth but that addition of the amine served to improve deactivation in some manner. While the exact role of the amine in this system remains unclear, ultimately these PCs were shown to produce PMMA with Đ as low as 1.27 and I* as high as 82%.

In another example, Liao reported the use of substituted BINOLs as PCs in O-ATRP. While some of these BINOLs would later be modified to generate heteroatom-doped PAHs (see , ), they were also shown to be effective O-ATRP catalysts for the synthesis of PMMA with Mn = 10.9–48.5 kDa, Đ = 1.2–1.6, and I* = 12% to 76%.

In 2019, Wang and co-workers studied a series of dyes in an effort to identify new catalysts for O-ATRP. Ultimately, several derivatives of quinacridone, indigo, and diketopyrrolopyrrole were shown to mediate O-ATRP. However, these PCs generally exhibited poor control in these polymerizations, producing polymer with Đ > 1.5 and I* < 90%. In part, this poor performance might be attributable to the poor electrochemical reversibility of the catalysts, which could inhibit their radical cations from successfully performing deactivation during O-ATRP. Later in 2020, Yang, He, and Jiang also explored substituted diketopyrrolopyrroles as O-ATRP catalysts. However, the polymers produced from MMA and styrene using these PCs were not characterized. Therefore, conclusions regarding the polymerization performance of these PCs cannot be made.

Also in 2020, Lei disclosed the use of triarylsulfonium hexafluorophosphate salts as PCs in the polymerization of MMA. In general, these PCs exhibited moderate polymerization control. The PMMA produced featured Mn ranging from 4.8 to 17.0 kDa, Đ = 1.3–1.6, and I* generally around 100%. Interestingly, the choice of solvent with this catalyst system was particularly important for maintaining a controlled polymerization, with less polar solvent systems resulting in the best performance.

While numerous PCs with varying properties have been reported for use in O-ATRP, one common downfall of many PCs is their ease of synthesis. Often, several-step syntheses are necessary to obtain successful O-ATRP catalysts, creating a barrier to their use. Addressing this issue, work reported by Yang showed that simple benzaldehyde derivatives could effectively mediate O-ATRP under the appropriate conditions. Similarly to benzotriazoles and benzothiadiazoles, these benzaldehydes showed improved polymerization control when used in the presence of sacrificial amines. However, because the polymerization proceeded also in the absence of amines, albeit with poor control (Đ > 1.5), it seems these benzaldehydes can successfully perform activation, whereas the role of the amines is to mediate deactivation of the polymer.

Finally, in 2020, riboflavin derivatives were shown to mediate the O-ATRP of methacrylates. In this report, riboflavin was functionalized to include pendant bromoisobutyrate groups, allowing it to function as both the PC and the initiator in O-ATRP (Figure 13). In the polymerization of hydroxyethyl methacrylate (HEMA), this catalyst/initiator system gave good control, yielding poly(HEMA) with Mn = 38.8 kDa, Đ = 1.2, and I% ~ 100%. Similar polymerization control was obtained in the polymerization of poly(ethylene glycol) methyl ether meth-acrylate (Mn = 116 kDa, Đ = 1.4, I* ~ 100%), albeit with slightly higher dispersity than in the polymerization of HEMA.


A bifunctional PC and initiator for O-ATRP based on riboflavin.

3.3.8. Reductive Quenching Photocatalysts.

While the majority of the catalysts employed in O-ATRP are PCs that operate through oxidative quenching (i.e., reduction of the alkyl halide by PC* to generate PC•+, see , ), a number of PCs operating through reductive quenching pathways have also been utilized. The first example of this approach came from Zhang and Cheng, who used fluorescein (FL, Figure 8) in the presence of triethyl amine to drive the polymerization of MMA. Unlike previous examples where an uncontrolled polymerization would proceed to high conversion in the absence of the amine,, the catalyst in this report appeared to proceed through a true reductive quenching mechanism. That is, in the absence of added amine, only a small degree of monomer conversion (2.6%) was obtained, indicating ineffective activation from PC*.

In the polymerization of MMA, fluorescein showed moderate polymerization control. In general, PMMA with Mn ranging from about 20 to 60 kDa was produced, with Đ from 1.3 to 1.6 and I* generally below 50%. Interestingly, similar polymerization control was obtained with styrene, although a monomer conversion of only 10% was achieved in this case. Nevertheless, experiments cycling the light source “on” and “off” showed monomer conversion was directly tied to irradiation of the reaction and that the polymer Mn and Đ remained constant during “off” periods. Further, MALDI-TOF analysis of the polymers produced by this method showed retention of the Br chain-end group in the polymerization of styrene, suggesting this polymerization proceeds through an ATRP mechanism. However, evidence was also found for significant loss of this chain-end functionality, as chain-extensions of PMMA with MMA and styrene showed a large quantity of unfunctionalized polymer remaining after the reaction. Therefore, while this report showed that O-ATRP could proceed in the presence of a reductive quenching PC, this method provided only moderate control over the polymerization.

Soon after this report, Yilmaz and Yagci also investigated reductive quenching PCs in O-ATRP. In addition to FL, eosin Y (EY) and erythrosin B were also studied and generally provided the best polymerization control in the O-ATRP of MMA. When EY was used as the catalyst, PMMA was obtained with Mn = 8.7–22.1 kDa and Đ = 1.3–1.9. Instead, erythrosin B gave PMMA with Mn = 13.7–90.0 kDa and Đ = 1.2–2.5. In addition, Wei and Chen showed eosin Y and rhodamine B could be grafted to cellulose to create recyclable PCs that maintain moderate control in the polymerization of MMA (Mn = 17.2–119 kDa, Đ = 1.3–1.6, I* = 3% to 25%). Finally, Zhang and co-workers demonstrated the successful O-ATRP of MMA in the presence of oxygen using fluorescein, paving the way to simplify experimental setups using reductive quenching PCs.

Traditional photoinitiators have also been employed as catalysts in O-ATRP, including thioxanthone, 2-isopropylthiox-anthone, benzophenone, and camphorquinone. In each case, moderate polymerization control was observed, with the polymers produced displaying Mn ranging from 4.8 to 17.0 kDa and Đ = 1.3–2.0. However, when the light source was cycled “on” and “off,” a small degree of monomer conversion was still observed during “off” periods. While the authors attributed this conversion to inefficient deactivation, the cause of this phenomenon remains unknown. Later in 2018, Yi showed that a substituted benzophenone derivative, functionalized to enable its solubility in water, could also mediate the O-ATRP of acrylamide, producing poly(acrylamide) with Mn = 2.7–37.5 kDa, Đ = 1.4–1.5, and I* = 67% to 97%.

Finally, Chmielarz and co-workers reported on the use of riboflavin as a reductive quenching PC in the presence of ascorbic acid as a sacrificial electron donor. Using this combination, the O-ATRP of a PEG methacrylate was undertaken using bromoisobutyrate-functionalized lignin as an initiator to create star-shaped polymers and copolymers.

3.4. Investigations of the O-ATRP Mechanism

Since the inception of O-ATRP in 2014, several investigations have sought to better understand the mechanism of this polymerization method. The primary focus of this work has often been understanding O-ATRP mediated by oxidative quenching PCs. As such, the following sections discuss each step of the oxidative quenching mechanism and are primarily organized according to the order of those mechanistic steps.

3.4.1. Photoexcitation and Photophysical Processes.

To truly understand the photophysics of O-ATRP PCs and the factors influencing their properties, it is important first to become familiar with photophysical processes that occur upon the absorption of light by a molecule. Here, we will provide a brief introduction to these processes as a foundation for subsequent discussions of PC photophysics. For further information on these topics, we refer the reader to other, more thorough resources on photochemistry.,,-

When a molecule absorbs light, it can undergo a number of photophysical processes that are often represented on a Jablonski diagram (Figure 14), the first being photoexcitation from a ground state to an excited state. In organic molecules, the ground state is often a singlet state (S0), and photoexcitation occurs to higher energy singlet states (Sn, n ≥ 1). In most cases, rapid photophysical processes such as vibrational relaxation and internal conversion cause relaxation to the lowest energy excited state (S1), so most of the photochemistry of interest occurs from this state. For example, in the absence of a quencher, a species that reacts with the excited state, the PC may undergo fluorescence (radiative relaxation) or internal conversion followed by vibrational relaxation (nonradiative relaxation) to transition from S1 to S0. Instead, if a suitable quencher is present, the PC may react with the quencher to initiate a reaction. This process can occur through either electron transfer or energy transfer, although the requirements for each mechanism differ. Further, reactivity from a singlet excited state can be limited by the short lifetime of the singlet excited state, typically picoseconds to nanoseconds. Because bimolecular collisions in solution typically require at least a few nanoseconds to occur,, a short excited-state lifetime can sometimes lead to relaxation prior to reaction with a substrate.


A general Jablonski diagram depicting common photophysical processes for organic molecules.

To overcome this limitation, some PCs can access longer-lived triplet excited states (Tn) through intersystem crossing (ISC), where the spin of the excited electron flips. During this process, the PC will transition from S1 to Tn (n ≥ 1), followed by rapid nonradiative decay to access T1. Because a spin-flip is quantum mechanically forbidden, relaxation of T1 to S0 is more challenging than S1 to S0, lengthening the lifetime of the triplet excited state, typically microseconds or longer. However, because the same is true for the transition from S1 to Tn, accessing reactivity from the triplet manifold is challenging and requires suitable catalyst design.

To this end, initial investigations of dihydrophenazine PCs yielded some insight into how catalyst structure can impact the triplet yields of these catalysts. During early work with this catalyst family, it was discovered that the identity of the N-aryl substituents could greatly influence the catalyst’s performance in O-ATRP. Specifically, catalysts with electron withdrawing groups (EWGs) consistently exhibited superior polymerization control (lower Đ) than those with electron donating groups (EDGs). Through computational investigations, it was discovered that PCs with EDGs also exhibited spatially separated singly occupied molecular orbitals (SOMOs) in 3PC*, whereas PCs with EDGs did not (Figure 9). In turn, it was hypothesized that the presence of EWGs led to ICT in PC* from the phenazine core to the N-aryl substituent, which could facilitate ISC and improve the triplet yield of the catalyst. In O-ATRP, the increased formation of 3PC* might improve activation, leading to better polymerization control as observed with some catalysts. To test this hypothesis, new PCs were targeted with 1-naphthyl and 2-naphthyl N-aryl substituents, which were also predicted by DFT to feature ICT. When employed in O-ATRP, these catalysts exhibited good polymerization control (Đ < 1.2), supporting the importance of this property for effective catalysis.

While this work provided useful guiding principles for future catalyst design, it relied primarily on computational evidence to show how certain N-aryl functionalities could impact PC photophysics. In later work, experimental evidence was found to support these theoretical insights, providing a stronger basis for conclusions drawn from this data. Because catalysts with ICT were predicted to feature polar excited states, due to localization of electron density on the N-aryl substituent, it was proposed that their emission would be susceptible to solvent polarity. In essence, increasing the polarity of the solvent might stabilize the excited state, leading to a red-shift in the emission of the catalyst. Indeed, when the emission of various phenazines was compared, PCs with ICT showed significant red-shifting of their emission in more polar solvents, while the emission of a PC without ICT remained essentially unchanged. Additionally, the fluorescence spectra of PCs with ICT showed broad, featureless peaks consistent with a charge transfer excited state, further supporting the conclusions of previous computational investigations. As such, the remaining question became exactly how ICT impacts PC photophysics to improve catalysis.

Motivated by this question, Damrauer and co-workers characterized the photophysics of two phenoxazine PCs: PhenO-1N-BiPh and the analogous N-phenyl derivative (PhenO-Ph-BiPh). Through these studies, the authors were ultimately able to propose an energy level diagram mapping the relaxation pathways of these two catalysts (Figure 15), yielding insight into the effect of the N-aryl substituent on photophysical relaxation processes. In the case of PhenO-Ph-BiPh, photoexcitation of the catalyst leads to the formation of a Franck–Condon singlet state (SFC), which rapidly relaxes to a singlet charge transfer state localized on a biphenyl core substituent (SCT-BiPh). While this charge transfer state enables ISC to an analogous triplet state (TCT-BiPh), this process is in competition with efficient fluorescence from SCT-BiPh to S0, making the quantum yield of ISC low (Φ = 0.11).


Energy diagrams demonstrating the impact of naphthyl N-aryl substituents on intersystem crossing in phenoxazine PCs dissolved in N,N-dimethylacetamide.

For PhenO-1N-BiPh, similar behavior is observed upon photoexcitation, ultimately leading to the formation of a similar SCT-BiPh state. However, the presence of the 1-naphthyl substituent in this catalyst leads to the formation of new intermediate states between SCT-BiPh and S0 that significantly impact subsequent relaxation processes. Although the final triplet excited state is similar to that in PhenO-Ph-BiPh (TCT-BiPh), an intermediate singlet state localized on the naphthyl substituent (SCT-Naph) provides more efficient ISC and greater yield of the triplet state (Φ = 0.91). The exact pathway through which this process occurs remains unknown, but it is hypothesized to occur either through direct relaxation of SCT-Naph to TCT-BiPh or through an intermediate dark state (unobservable) localized on the naphthyl ring (TCT-Naph). Regardless of the pathway, the effect is the same: ICT to the naphthyl substituent enables more efficient ISC, increasing the [3PC*] available to engage in catalysis during O-ATRP.

In later work, Damrauer and co-workers expanded on these studies to investigate the effect of naphthyl connectivity on PC photophysics in these phenoxazines. Since phenoxazines with 1-naphthyl and 2-naphthyl N-aryl groups had previously been reported for O-ATRP,, this investigation sought to understand whether the location of naphthyl connectivity could impact important catalyst properties. Again, the photophysical relaxation processes of various catalysts were characterized, and it was revealed that the naphthyl connectivity could have a small impact on the energy of the SCT-BiPh state. In the case of PhenO-2N-BiPh (a 2-naphthyl phenoxazine PC), the change in naphthyl connectivity results in a slight destabilization of SCT-Naph, leading to an equilibrium between SCT-Naph and SCT-BiPh. Consequently, ISC becomes less competitive and the yield of 3PC* becomes significantly lower (Φ = 0.54). In addition, the reorganization energy for charge transfer from the N-aryl substituent to the PC core during ISC was found to be roughly 10% larger for 2-naphthyl versus 1-naphthyl substituents, presumably due to the larger donor–acceptor distance between the PC core and the 2-naphthyl group. As a result, PCs with 2-naphthyl groups exhibit slower intersystem crossing (smaller kISC) than those with 1-naphthyl groups. Importantly, both of these observations, excited-state energies and reorganization energy, could impact catalysis where formation of 3PC* is critical to the success of the reaction.

Further expanding on these investigations, it was shown that these principles could be applied to other catalyst families to design high triplet yield PCs. Within the phenothiazine family, a series of catalysts was designed with various N-aryl functionalities intended to stabilize the SCT state to increasing degrees. Through characterization of the photophysics of these PCs, it was shown that stabilization of this state could be used to increase the yield of 3PC* up to 96%. However, excessive stabilization, through introduction of strong EWGs, could also bypass the triplet manifold and result in efficient nonradiative decay to S0.

While the work discussed to this point yielded important insights into the photophysics of common O-ATRP catalysts, others have focused their investigations on the impact of external factors on photoexcitation and photoredox catalysis in O-ATRP. For example, Ryan et al. probed the impact of light intensity on polymerization control in the presence of perylene and PhenO-1N-BiPh. Interestingly, lowering the intensity of the light source resulted in a gradual decrease in polymerization control. It was hypothesized that this observation was due to decreasing efficiency of activation, which ultimately results in insufficient buildup of the PC•+ deactivating species. As a result, deactivation becomes inefficient and polymerization control is lost at low light intensities. More interesting, however, was the discovery that tolerance to low light intensity could be designed into the catalyst, as PhenO-1N-BiPh operated effectively at lower light intensities than perylene. While the exact reason for this superior performance has not been investigated, one can hypothesize that the stronger excited-state reduction potential [E°(PC•+/3PC*)] and higher molar absorptivity of PhenO-1N-BiPh are likely beneficial under these conditions.

An investigation by Hawker and co-workers probed various controlled radical polymerizations in the dark to understand how the polymerization process was impacted by manipulation of the light source (i.e., turning it on or off). In this report, the advantage of photoredox catalyzed ATRP over traditional, Cu mediated photoATRP was demonstrated by tracking monomer conversion in situ during periods of irradiation and darkness. When irradiation was ceased, Cu mediated photoATRP still showed slow monomer conversion in the dark, whereas O-ATRP catalyzed by PhenS-Ph stopped immediately in the dark. In addition, when the light source was turned on again, Cu catalyzed photoATRP showed nonlinear, gradually increasing kinetics, whereas O-ATRP with PhenS-Ph immediately exhibited linear pseudo-first-order kinetics.

To understand these differences, it is important to understand how the mechanisms of photoATRP and O-ATRP differ. In photoATRP, irradiation of the reaction solution enables conversion of CuII to CuI, a long-lived species capable of mediating ATRP in the absence of light. Instead, irradiation in O-ATRP generates a short-lived PC excited state, which rapidly relaxes in the dark to a ground state that is incapable of mediating O-ATRP on its own. This rapid relaxation is supported by the short excited-state lifetimes of O-ATRP PCs, as well as kinetic modeling performed by Guo and Luo. As a result, O-ATRP offers precise temporal control over the polymerization, while photoATRP is generally less responsive on shorter time scales.

3.4.2. Activation.

One longstanding question in O-ATRP has been the nature of the catalyst excited state responsible for catalysis. Specifically, is it the singlet state (1PC*) or the triplet state (3PC*) that is most relevant (Figure 16)? Indeed, arguments can be made for each one. While 3PC* is likely longer lived, making it more likely to engage in bimolecular reactions in solution, it is generally less reducing than 1PC* due to photophysical relaxation processes. By the same token, 1PC* is more reducing than 3PC*, which increases the driving force for activation, but it may be too short-lived to efficiently undergo bimolecular reactions. Therefore, which of these properties is most important in O-ATRP?


Advantages and disadvantages associated with 1PC* and 3PC* for activation during O-ATRP.

In 2016, Jockusch and Yagci attempted to answer this question through investigation of methyl phenothiazine (PhenS-Me). In this work, the rate of electron transfer (ET) from 1PC* to several alkyl halides was measured by fluorescence spectroscopy, and that from 3PC* was measured using transient absorption spectroscopy. It was found that the rates of ET were generally greater for 1PC*, a discovery that is unsurprising given the greater reduction potential of 1PC* relative to 3PC*. However, PhenS-Me exhibited significant ISC to the triplet state (Φ ~ 0.6), suggesting the yield and lifetime of 3PC* could counteract its lower rate of ET. As such, these authors concluded activation in O-ATRP with PhenS-Me most likely occurs from 3PC*, with minor contributions from 1PC*.

In another study, Orr-Ewing and co-workers also investigated activation during O-ATRP, although this time using picosecond transient absorption spectroscopy. In this investigation, ET between methyl-2-bromopropionate and PhenN-PhCF3 or the N-phenyl analogue (PhenN-Ph) was targeted, because PhenN-PhCF3 was previously proposed to operate via 3PC*, while PhenN-Ph was proposed to operate via 1PC*. By varying the polarity of the solvent, the authors showed that the rate of ET could be influenced for PhenN-PhCF3 but not PhenN-Ph, consistent with previous suggestions that the ICT nature of the PhenN-PhCF3 excited state made it susceptible to solvent polarity. More importantly, however, these studies revealed PhenN-Ph exhibits faster ET than PhenN-PhCF3 and that both catalysts can perform activation from 1PC*. In turn, this result led the authors to conclude that the most successful catalysts for O-ATRP should feature short 1PC* lifetimes, low yields of ISC, and slow ET to minimize the concentration of radicals in solution.

In later work, this investigation was expanded to include PhenN-2N, another common catalyst in O-ATRP. Again, the results of these investigations suggested ET occurs primarily from 1PC* to the alkyl bromide, contrary to previous suggestions that a triplet excited state may be catalytically active. Further, it was found that ET from PhenN-2N was generally slower than from PhenN-PhCF3 or PhenN-Ph, again suggesting slower rates of ET may be most beneficial for successful O-ATRP (because PhenN-2N is one of the top performing catalysts for this method). These conclusions were further supported in 2021, when these investigations were expanded to a series of nine PCs spanning three PC families: dihydrophenazines, phenoxazines, and phenothiazines.

It is important to note, however, that these investigations did not consider the effect of deactivation in O-ATRP. Indeed, in the absence of deactivation, slow, inefficient ET would be beneficial to suppress the concentration of propagating radical, which is necessary to limit irreversible termination reactions in O-ATRP. However, the suppression of radicals in O-ATRP is also achieved through the deactivation step of the mechanism. As such, it may be possible to design catalysts with fast and efficient ET that also effectively mediate O-ATRP, as long as those catalysts are capable of performing effective deactivation to control the concentration of propagating radicals. In light of this consideration, it is possible that the superior performance of PCs with ICT such as PhenN-PhCF3 or PhenN-2N may be attributable to their ability to deactivate alkyl radicals rather than their photophysical properties. Instead, PCs such as PhenN-Ph may be less successful in O-ATRP due to the low oxidation potentials of their radical cations [E°(PC•+/PC)], which will impede deactivation.

One other important takeaway from the studies by Orr-Ewing and co-workers is that effective catalysts for O-ATRP, PhenN-PhCF3 and PhenN-2N, exhibited reactivity primarily from a singlet excited state with charge transfer character., Although these studies call into question the role of 3PC* in activation, they do provide support for the importance of PCs exhibiting ICT in the excited state. As such, this property remains an important design principle for the development of new PCs for O-ATRP.

To further probe which excited state is relevant for O-ATRP, Damrauer and co-workers investigated the activation of an alkyl bromide by four phenoxazines with varying N-aryl and core substituents. In this study, ET rate constants for both 1PC* and 3PC* were measured, as well as reaction quantum yields for each excited state. As a result, it was observed that both 1PC* and 3PC* can contribute significantly to activation during O-ATRP, but which one contributes most can depend on a number of factors. Consistent with investigations by Orr-Ewing,- the driving force for electron transfer (ΔG°ET) is extremely influential and favors activation from 1PC* because this state tends to be more reducing than 3PC*. However, the lifetime of the excited state and the yield of ISC (ΦISC) are also important factors to consider. At low concentrations of quencher (i.e., the alkyl halide) where bimolecular reactions are less likely, when ΦISC is high, or if the lifetime of 3PC* is long, reactivity from 3PC* can contribute significantly to activation. As such, this work highlights the importance of considering both 1PC* and 3PC* in O-ATRP, as both states can contribute to catalysis. Further, it is worth noting that the relative contributions of 1PC* and 3PC* likely vary from one catalyst family to another, between catalysts within the same family, and even for the same catalyst under different reaction conditions. For example, because dihydrophenazines are typically more reducing in the excited state, feature shorter triplet lifetimes, and lower ΦISC than phenoxazines, reactivity from 1PC* may be more significant for dihydrophenazines than other catalyst families. In addition, performing polymerizations at low catalyst or initiator concentrations may favor reactivity from 3PC* because lowering the concentrations of these species will make bimolecular reactions with a short-lived 1PC* more challenging. Another consideration is that O-ATRP reactions are typically performed at high monomer concentrations. During the course of polymerization, the solution viscosity significantly increases, impeding diffusion. Therefore, long-lived excited-state PCs may become more important later in the polymerization relative to at the onset of polymerization.

Regardless of the relative contributions of 1PC* and 3PC*, Matyjaszewski and co-workers showed that activation in O-ATRP likely occurs through a dissociative, outer-sphere electron transfer mechanism, which is not typically observed in metal-catalyzed ATRP. Moreover, activation in O-ATRP is generally very fast, approaching diffusion limited kinetics for many PCs. This observation has since been supported by others-, and seems to be a general trend for many O-ATRP catalysts. However, because of the short lifetimes of many PCs, the efficiency of activation tends to be low, resulting in an effective rate of activation similar to that observed in metal catalyzed ATRP.

Finally, Pearson et al. used computational methods to investigate how reorganization of different PC cores during electron transfer can contribute to activation rates during O-ATRP (Figure 17). In this work, three catalysts with N-phenyl substituents were investigated: PhenO-Ph, PhenN-Ph, and PhenS-Ph. In each case, the reduction potentials and reorganization energies relevant to activation (3PC* to PC•+) were computed by DFT, revealing that all three catalysts feature a similar driving force for activation [E°(PC•+/3PC*) ~ −2 V vs SCE]. However, because PhenS-Ph contains a larger S atom in its core, this catalyst exhibits a significantly larger reorganization energy than PhenO-Ph. As a result, it was predicted PhenS-Ph would exhibit slower activation than PhenO-Ph, suggesting even small structural changes in the PC core can significantly influence catalysis.


Impact of PC core on reorganization energy (λ) during electron transfer. Key: white = hydrogen, gray = carbon, blue = nitrogen, red = oxygen, yellow = sulfur.

3.4.3. Deactivation.

One of the most important steps in the mechanism of O-ATRP (and more broadly ATRP as well as all controlled radical polymerizations) is deactivation. In any radical polymerization method, bimolecular radical termination reactions will always be present to some degree, necessitating a method to inhibit these side reactions and maintain control of the polymerization. In O-ATRP, this method involves reversible deactivation of the propagating radical through installation of a halide on the polymer chain-end. In effect, this process reduces the concentration of radicals to prevent radical-based side reactions while enabling the polymer to be reactivated for future chain growth.

Despite the importance of this mechanistic step, deactivation remains relatively understudied in comparison to photoexcitation and activation in O-ATRP. In one of the most comprehensive studies of this step, Matyjaszewski and co-workers probed deactivation through the addition of halide salts to the polymerization of MMA in the presence of PhenS-Ph. Hypothesizing that deactivation could be mediated by the radical cation ion pair (PC•+X−, X = Br or Cl), the authors proposed that the addition of halide salts to the polymerization should encourage the formation of this ion pair. The result would then be more effective deactivation, observable through the lowering of the polymerization rate and improvements in polymerization control (i.e., lower Đ, I* closer to 100%).

When this experiment was performed in O-ATRP using EBP as the initiator, a slight decrease in the rate of the polymerization was observed but without any significant improvement in polymerization control, potentially indicating unanticipated complexities in this system. However, a noticeable effect was observed when the initiator was changed to ethyl α-chlorophenylacetate (EClP). In comparison to polymerizations without added Br−, those with additional Br− showed significant improvements in I* (75% with added Br− versus 9% without it). In addition to highlighting the importance of Br− for deactivation, this experiment suggests deactivation is particularly ineffective in the presence of Cl−. Further supporting this conclusion, when the authors performed additional polymerizations using EClP, they discovered that PhenS-Ph could activate this initiator because 55% monomer conversion was observed in 4 h (compared to 15% in the same time without catalyst present). However, the polymerization was completely uncontrolled (Đ = 3.4, I* = 36%), suggesting deactivation with Cl− had been ineffective.

To gain deeper insight into deactivation in O-ATRP, Matyjaszewski and co-workers then employed computational chemistry to compute the thermodynamic feasibility of five possible deactivation mechanisms (Figure 18): two proceeding through an outer-sphere electron transfer [OSET (1) and OSET (2)], one through an inner-sphere electron transfer (ISET), one through dissociative electron transfer (DET), and another by termolecular associative electron transfer [AET (ter)]. The results of these calculations showed that the AET (ter) mechanism is most favorable, but it is important to consider that these calculations only yield insight into the thermodynamics of each mechanism. To understand kinetic contributions, activation energies computed by DFT were used to estimate rate constants for each deactivation mechanism. The results of these calculations showed that of the five proposed mechanisms, only OSET (2) and AET (ter) could outcompete the rate of termination in the polymerization of MMA. However, the estimated rate of deactivation through AET (ter) was nearly 2 orders of magnitude larger than through OSET (2), again suggesting deactivation may occur through this termolecular mechanism.


Possible mechanisms of deactivation proposed by Matyjaszewski in 2016 (left) and structures of relevant intermediates (right).

Later in 2017, Lim et al. probed the impact of solvent polarity on ion pairing in PC•+Br− and the effect of manipulating this variable on polymerization control in O-ATRP. Hypothesizing PC•+Br− to be the deactivator in O-ATRP, these authors reasoned that lowering the polarity of the reaction solution would favor formation of this ion pair, in turn increasing the rate of deactivation through increasing [PC•+Br−]. To probe the impact of solution polarity, DFT calculations were used to calculate the association free energy (ΔG°assoc) for a phenazine radical cation (PhenN-1N) and Br−. Consistent with their expectations, these calculations predicted that ΔG°assoc would become increasingly exergonic with decreasing solvent polarity, suggesting this variable could be tuned to manipulate deactivation in O-ATRP. As such, a series of polymerizations was performed with varying ratios of N,N-dimethylacetamide (DMAc) and tetrahydrofuran (THF) as the solvent system. Interestingly, the addition of 25% THF improved polymerization control, suggesting decreasing solvent polarity could be beneficial to polymerization outcomes in O-ATRP. Expanding on this work, later reports showed that these results could be generalized to a series of dihydrophenazine PCs and solvent systems, and that tuning solvent polarity could also be used to lower catalyst loadings in O-ATRP (see , ).

In work by Guo and Luo, kinetic modeling was employed to understand how changes in activation can impact deactivation during O-ATRP. While one might initially consider these steps to be independent, it is important to remember that the buildup of deactivator during early polymerization times is directly dependent upon activation. As such, activation and deactivation are intimately intertwined, especially early in the reaction (or rather at early times after irradiation is started). Through kinetic modeling, Guo and Luo demonstrated this dependence on activation by showing that [PC•+] must surpass a threshold before deactivation can become effective. To reach this threshold rapidly, fast and effective activation is necessary. To manipulate this process, control of the reactor light intensity can be crucial, as the authors showed increasing light intensity could improve polymerization control by increasing the rate of activation and therefore the rate of PC•+ buildup. Importantly, these results are consistent with previous reports on the effect of light intensity in O-ATRP, which suggested the same link between light intensity, the rate of activation, and efficiency of deactivation.

Finally, work by Corbin et al. probed deactivation in O-ATRP by synthesis and investigation of the PC radical cations hypothesized to mediate this process. Through a model reaction using azobisisobutyronitrile (AIBN) to generate alkyl radicals, evidence was found supporting the ability of a phenazine radical cation to deactivate radicals in the presence of BR−. Through further investigation of this model reaction, it was also discovered that the oxidation potential of the halide [E°(X•/X−), X = Br or Cl] as well as the radical cation [E°(PC•+/PC)] impact the rate of deactivation. By measuring ΔG°assoc for the formation of PC•+PF6− ion pairs, it was found that catalyst structure has a minimal influence on ion pair strength, whereas tuning solvent polarity greatly impacts this property. As such, changing the solvent in O-ATRP may be the most effective approach to manipulating deactivation through ion pairing. Finally, when a radical cation was employed as a reagent in O-ATRP, this strategy yielded improved control in the polymerization of acrylates, further supporting the role of PC•+ in deactivation during O-ATRP.

3.4.4. Side Reactions.

Perhaps the most obvious side reactions in O-ATRP are the termination reactions inherent to any radical polymerization method. Often, termination occurs through one of two common mechanisms: radical coupling (also called combination) or disproportionation. In essence, radical coupling involves the reaction of two propagating radicals, which couple to each other to irreversibly form a C─C bond. Instead, disproportionation involves hydrogen atom abstraction from the carbon beta to one propagating radical by a second radical in solution. Because these modes of termination are general to all radical polymerization methods and have been discussed in detail by others,,, they will not be discussed further in this text. Instead, this discussion will focus on side reactions more unique to O-ATRP.

In 2016, Matyjaszewski and co-workers investigated the ability of the alkyl halide initiator to undergo direct activation under irradiation with UV light. Their results showed that both EBP and EClP could initiate uncontrolled polymerizations in the absence of a PC, suggesting direct photolysis of the initiator may be a major side reaction in O-ATRP. In the case of EBP, however, addition of a PC to the reaction solution slowed monomer conversion and improved polymerization control, suggesting the presence of an appropriate catalyst can limit this side reaction. In addition, the use of visible light irradiation, which is likely not absorbed by alkyl halides, could also assist in minimizing these side reactions.

Further evidence supporting this side reaction was later presented by Guo and Luo, who used kinetic modeling to show that a small quantity of polymer chains can be generated in O-ATRP through direct photolysis of the initiator. Importantly, their work showed this side reaction can become problematic at low catalyst loadings, where the [PC•+] remains limited and deactivation cannot control the propagation of these undesired radicals. As such, a threshold exists, one that likely varies depending on the catalyst and reaction conditions, below which the catalyst is in too low of a concentration to effectively mediate O-ATRP.

Another side reaction in O-ATRP involves substitution of the PC by alkyl radicals during the polymerization., In initial reports on dihydrophenazine catalysts, it was observed that successful PCs could produce PMMA with good dispersity (Ð < 1.2) but consistently low initiator efficiency (I* = 60% to 80%)., Even more interestingly, when one of these dihydrophenazines was further functionalized by installation of aryl core substituents, these new PCs consistently produced polymer with I* ~ 100%. In turn, these observations motivated the hypothesis that certain O-ATRP catalysts could undergo substitution by the initiator during a polymerization, which would ultimately result in low I*. The ability of one dihydrophenazine PC (PhenN-2N) to undergo substitution by an alkyl bromide was confirmed by McCarthy et al., which ultimately enabled good control in the polymerization of acrylate monomers. In addition to being isolatable, this substituted catalyst could be prepared in situ prior to polymerization by preirradiation of a catalyst/initiator solution, followed by addition of the monomer and additional initiator to start the polymerization.

After this report, a systematic investigation of this side reaction was performed by Corbin et al., revealing that dihydrophenazine core-substitution can occur via a number of alkyl bromide initiators as well as the propagating polymer in O-ATRP. Importantly, substitution by the propagating polymer represents a new termination reaction that was previously unknown in O-ATRP. Through investigation of initiators with different sterics, it was discovered that tertiary bromides could undergo two additions to the catalyst core, whereas secondary bromides could undergo up to four additions to the catalyst core (Figure 19). Further, the use of core substituted catalysts in O-ATRP consistently produced polymers with I* ~ 100%, supporting the hypothesis that low I* with noncore substituted catalysts is due to an in situ core-substitution side reaction. A similar study was also reported by Zhao, Guo, and co-workers, which investigated the core substitution of dihydrophenazines using various alkyl bromides and benzyl bromide derivatives. Interestingly, this report showed unsubstituted phenoxazines and phenothiazines can undergo similar substitution-based side reactions, suggesting this reactivity is not unique to a single catalyst family.


Observed core-substitution of dihydrophenazine PCs by alkyl radicals such as those present during O-ATRP. Note that R-Br can be either an alkyl bromide initiator or the alkyl bromide chain-end group of a polymer. *Only the 2,7 isomer is depicted, although the 2,8 isomer is also formed.

Finally, in an investigation of deactivation in O-ATRP, several side reactions related to this step were probed and identified. Perhaps most significantly, it was discovered that PC•+ can react with DMAc, a common O-ATRP solvent, as well as halides such as Br− and Cl−. In both cases, PC•+ is reduced via single electron transfer to generate the neutral PC, a process that is accelerated in light and that likely impedes deactivation under the appropriate conditions. Because this process can generate halogen radicals (Br• and Cl•), the propensity for these radicals to engage in hydrogen atom abstraction under O-ATRP conditions was evaluated as well. While this possible side reaction was not ruled out, it did not appear to be greatly impacted by the identity of the halide in the presence of MMA. As such, because well controlled PMMA has been synthesized repeatedly in the presence of bromides, these results suggest this side reaction may not be significant in the O-ATRP of MMA.

3.4.5. Solvent Effects in O-ATRP.

Early in the development of O-ATRP, researchers noted that solvent choice could impact catalyst properties and the outcome of a polymerization. For example, dihydrophenazines exhibiting ICT in the excited state displayed solvatochromic emission from stabilization of the excited state in more polar solvents, suggesting the excited-state properties of these PCs are significantly impacted by changes in solvent polarity. In addition, computational work suggested lowering solvent polarity could encourage formation of the PC•+Br− ion pair, a species proposed to mediate deactivation in O-ATRP. Together, these results led researchers to believe that performing O-ATRP in lower polarity solvents could improve polymerization control by improving activation (through manipulation of PC*) and deactivation (by encouraging ion pairing in PC•+Br−). Indeed, when O-ATRP was performed with a mixture of DMAc and THF, improved polymerization control was obtained relative to using DMAc as the solvent alone. In a later study, this point was further demonstrated by investigation of O-ATRP in several solvent systems. Interestingly, catalysts with ICT in the excited state were most robust to changes in solvent polarity and ultimately performed best in less polar solvents, consistent with the hypotheses presented in previous work.

While these studies provided important insight into the importance of solvent choice in O-ATRP, the fundamental impacts of solvent polarity on PC properties remained poorly understood for some time. Later in 2020, McCarthy et al. performed a systematic investigation of solvent effects on the PhenN-2N catalyst, yielding deeper insight into this phenomenon. Through detailed investigation of PC photophysics in DMAc (more polar) and THF (less polar), these authors identified several important implications of lowering solvent polarity. These included a decrease in nonradiative decay from S1 to S0, an increase in intersystem crossing from the singlet to the triplet manifold, and an increase in the lifetimes of both the S1 and T1 excited states. Similar results have been reported by Orr-Ewing, and all three effects can be expected to improve activation during O-ATRP by increasing the concentration of excited states available for catalysis. In addition, the PC•+ of PhenN-2N was found to be more oxidizing [larger E°(PC•+/PC)] in less polar solvents, presumably due to destabilization of the cation. In turn, this property increases the driving force for deactivation, which should improve polymerization control in O-ATRP. As such, several synergistic effects exist upon lowering solvent polarity that will ultimately benefit O-ATRP. In both the polymerization of acrylates and the polymerization of MMA at low catalyst loadings, these solvent effects were evident.

Finally, in the investigation of radical cations and deactivation in O-ATRP, several solvent effects were identified. As predicted previously through computational studies, solvent polarity was shown to influence ion pairing in radical cation salts (PC•+PF6−) by increasing ΔG°assoc with decreasing polarity. In addition, it was found that several radical cations can oxidize O-ATRP solvents, such as DMAc and DMF. As such, to ensure stability of PC•+ during O-ATRP, the use of solvents with greater oxidation potentials [E°(S+/S), S = solvent] is beneficial. In this work, ethyl acetate was a suitable choice in which radical cation decomposition could be minimized.

One detail that should be noted is that most of the studies presented here have focused on the dihydrophenazine family of PCs. While these investigations can serve as useful guides to choosing solvents for O-ATRP, they may not be generalizable to other catalyst families. Moreover, there may be other solvent considerations that are not directly related to PC properties. For example, in some instances dichloromethane may be a poor solvent choice for O-ATRP because some PCs may be able to reduce this solvent [E° (CH2Cl2/CH2Cl2•−) = −2.2 to −2.5 V vs SCE], leading to unwanted initiation and termination reactions. In addition, the use of toluene as a solvent could be detrimental to polymerization control because the benzylic C─H bond in toluene is easily broken and the resulting benzylic radical is stabilized by resonance. As a result, toluene can serve as a chain-transfer agent in radical polymerizations, again causing unwanted initiation and termination of polymer chains. Moving forward, further investigations elucidating the impact of various solvents in O-ATRP will undoubtably be useful to the development of the field. In particular, studies revealing solvent effects for new catalyst families (i.e., other than dihydrophenazines) or that develop a general guide to understanding solvent effects in O-ATRP could significantly impact the field.

3.4.6. Mechanistic Insights in O-ATRP by Reductive Quenching.

Compared to O-ATRP by oxidative quenching, fewer studies exist on the mechanism of O-ATRP using reductive quenching catalysts. In 2018, Luo and co-workers used kinetic modeling to probe this process, the results of which support the importance of the sacrificial electron donor in these methods. For example, their modeling suggested that in the absence of a donor, no polymerization would occur. This result was consistent with experiments, which also showed no polymerization in the presence of EY without a sacrificial electron donor also present. In addition, it has been proposed that deactivation in these methods is mediated by the radical cation of the sacrificial electron donor, often an amine radical cation. In this work, kinetic modeling predicted that the concentration of the donor would remain constant throughout the polymerization, consistent with the donor radical cation being reduced during deactivation to regenerate the original donor species.

While these studies are insightful and can guide the development of this method, it is important to note that they are predicated on a hypothesized mechanism and do not consider alternative mechanisms. As an example, since a single mechanism of deactivation involving the donor radical cation was considered, this modeling cannot reveal the existence other possible deactivating species. As such, further investigation of the mechanism of O-ATRP using reductive quenching PCs, including experimental kinetics to validate this model, is warranted and would likely benefit future development in this field.

4. MONOMERS POLYMERIZED BY O-ATRP

4.1. Methacrylates

By far the most common monomers in O-ATRP have been methacrylates (Figure 20). First reported in the seminal works by Miyake and Theriot and Hawker, MMA has been the monomer of choice for most new O-ATRP methods. Often, it can be polymerized to molecular weights ranging from 1 to 20 kDa with Ð < 1.2 and I* ~ 100%. Hawker showed PhenS-Ph could also polymerize dimethylaminoethyl methacrylate with good control (Mn = 8.8 kDa, Ð = 1.1), highlighting for the first time an advantage of O-ATRP over traditional metal-catalyzed ATRP methods. Because this monomer can coordinate to metal catalysts and alter their catalytic properties, the use of metal-free catalysts in O-ATRP was crucial for its successful polymerization. In addition, both Hawker and Miyake showed other methacrylates could be polymerized in the synthesis of block copolymers with PMMA. In Hawker’s case, benzyl methacrylate (BnMA) was used to synthesize PMMA-b-PBnMA, with Mn = 25.9 kDa and Ð = 1.3. Instead, Miyake and Theriot synthesized PMMA-b-PBMA (BMA = n-butyl methacrylate) by converting PMMA (Mn = 72.9 kDa, Ð = 1.3) to a block copolymer with Mn = 523 kDa and Ð = 2.6.


Structures of methacrylate monomers polymerized by O-ATRP.

In 2016, the monomer scope of O-ATRP was expanded to several other methacrylates using two dihydrophenazine catalysts. These monomers included trimethylsilyl hydroxyethyl methacrylate (Mn = 20.0 kDa, Ð = 1.3, I* = 86%), diethyleneglycol methyl ether methacrylate (Mn = 21.3 kDa, Ð = 1.4, I* = 85%), and trifluoroethyl methacrylate (Mn = 54.7 kDa, Ð = 1.1, I* = 24%). While each of these methacrylates was polymerized with varying Ð and I*, highlighting differences in polymerization control, the broad array of functionalities within these monomers began demonstrating the excellent functional group tolerance enjoyed by O-ATRP.

Broadly, methacrylates with alkyl chains of various lengths have been well tolerated in O-ATRP. These include methyl methacrylate,, ethyl methacrylate, n-butyl methacrylate,, i-butyl methacrylate, t-butyl methacrylate, 2-ethylhexyl methacrylate, i-decyl methacrylate, and dodecyl methacrylate. Because numerous different catalysts have been employed to achieve this wide monomer scope, a direct comparison of polymerization control across monomers of increasing alkyl chain length is challenging. However, a sense of the effect this group has can be gained from work investigating the polymerization of various methacrylates in continuous flow using a single catalyst. In this case, increasing the chain length generally resulted in a loss of polymerization control, as observed through increasing Ð and gradually decreasing I* (Table 3). In part, this loss of polymerization control may be attributable to the increase in the rate of propagation (represented by kprop, Table 4), which makes propagation more challenging to control in the absence of sufficient deactivation. However, changes in monomer structure can also be expected to impact other polymerization conditions such as the reaction solution polarity. Because a number of catalytic properties (ex., PC photophysics, ion pairing, etc.) are susceptible to changes in solution polarity (see , ), it is difficult to understand exactly why increasing the length of the methacrylate alkyl chain decreases polymerization control. As such, further investigation of this phenomenon is necessary.

Table 3.

Impact of Methacrylate Alkyl Chain Length on Polymerization Control As Observed by Ramsey et al.

alkyl chainMn (kDa) Đ I* (%)ethyl7.61.294i-butyl9.91.2962-ethylhexyl16.01.498i-decyl26.51.256dodecyl18.51.583

Table 4.

Free Radical Propagation Rate Constants of Common Vinyl Monomers Calculated at 30°C

monomerAa × 10−6 (l mol−1 s−1)EAb (kJ mol−1)kpropc (M−1 s−1)refMethacrylate Monomersmethacrylic acid0.3816.1639 methyl methacrylate2.6722.4369 ethyl methacrylate4.0623.4377 n-butyl methacrylate3.7822.9428 i-butyl methacrylate2.6421.8463 2-ethylhexyl methacrylate1.8720.4571 i-decyl methacrylate2.1920.8571 dodecyl methacrylate2.5021.0602 cyclohexyl methacrylate4.8822.3701 3.7621.5742 benzyl methacrylate3.6121.5713 8.5023.2855 i-bornyl methacrylate4.2822.5568 glycidyl methacrylate4.4121.9743 6.0222.9682 hydroxyethyl methacrylate8.8821.91,500 2-hydroxypropyl methacrylate3.5120.8915 Acrylate Monomersmethyl acrylate16.617.714,800 n-butyl acrylate15.817.316,500 dodecyl acrylate17.917.021,100 Other Monomersacrylonitrile16.21,100 styrene43.032.5108 4.5026.0149 vinyl acetate10.019.83,870 14.720.73,990 1,3-butadiene80.535.757 chloroprene20.026.6522

In addition to these alkane-functionalized methacrylates, a number of methacrylates with heteroatom-containing aliphatic groups have been polymerized successfully. These monomers include methacrylic acid, 2-hydroxyethyl methacrylate, diethylene glycol methyl ether methacrylate, poly(ethylene glycol) methacrylate,, glycidyl methacrylate, and diethylamino ethyl methacrylate. A number of aromatic monomers have also been polymerized, including benzyl methacrylate, and furfuryl methacrylate,, although not as extensively as aliphatic methacrylates.

Exploiting the wide functional group tolerance of O-ATRP, a number of reports have disclosed the polymerization of highly functionalized monomers for various applications. In 2019, Ni and Niu reported the successful polymerization of several azide containing methacrylates with good polymerization control (Mn = 11.9–22.6 kDa, Ð = 1.2–1.3, I* = 70% to 100%). Using FT-IR, these authors demonstrated the presence of the desired azide functionality within the product polymer, supporting the tolerance of this method to these functional groups.

In other work, researchers disclosed the polymerization of 2-([4,6-dichloro-triazin-2-yl]oxy)ethyl methacrylate, pyrenyl methacrylate, and even fluorescein-o-methacrylate. In addition, several bifunctional monomers have been reported in O-ATRP, such as allyl methacrylate, methacrylate-based inimers (monomers that can also serve as initiators), and dimethacrylate monomers to achieve polymer cross-linking. Because of the complex polymer architectures achieved using such monomers, in particular inimers and dimethacrylates, polymerization control in these systems often cannot be evaluated.

Further highlighting the excellent functional group tolerance of O-ATRP, several research groups have reported the polymerization of metal-containing monomers using this method. In an example from the Hawker group in 2018, methacrylates functionalized using tethered Ir complexes were prepared and grafted to Si surfaces. By tuning the functionalities on each complex, the authors were then able to tune the emission of the resulting films, generating patterned films of various colors under UV irradiation. Similarly, Kong and co-workers disclosed the polymerization of ferrocenylmethyl methacrylate to generate ferrocene-containing polymers, which were ultimately employed in the detection of lung cancer DNA.

In other reports, the metal-free conditions enabled by O-ATRP have been exploited in the synthesis of polymers for metal-sensitive applications. One such application is in electronics, where residual metal contaminants within the polymer can lead to detrimental side reactions and undesirable material performance. For this reason, O-ATRP was chosen for the synthesis of poly(PEG) methacrylate lithium sulfonyl-(trifluoromethylsulfonyl)imide), a single-ion homopolymer electrolyte intended for battery applications. Another area where the elimination of metal contamination can be beneficial is in biological and medical applications. For this reason, O-ATRP was selected for the copolymerization of fluorescein O-methacrylate and sulfobetaine methacrylate to generate polymer-based drug delivery vehicles. Similarly, nanodiamonds were surface-functionalized with 2-methacryloyloxyethyl phosphorylcholine, a zwitterionic methacrylate, with the goal of generating new materials for biomedical applications. Importantly, each of these examples highlights the polymerization of monomers with ionic and other functionalities, further highlighting the excellent functional group tolerance enjoyed by O-ATRP.

Finally, Chu and Tang showed that several monomers derived from biomass could also be polymerized by O-ATRP. In this work, the authors synthesized methacrylate-based monomers from soybean oil, rosin acid, and furfural, three biomass feedstocks. Using PhenS-Ph, they showed these monomers could be polymerized with good to moderate control (Mn = 2.5–11.0 kDa, Ð = 1.1–1.4), again demonstrating the utility of this highly tolerant polymerization method.

4.2. Acrylates

Closely related to methacrylate monomers, acrylates (Figure 21) have generated significant interest in efforts to expand the monomer scope of O-ATRP. However, several challenges exist in polymerizing this monomer family. For example, acrylates typically exhibit kprop values roughly an order of magnitude larger than methacrylates (Table 4). As such, much more efficient deactivation is necessary to maintain polymerization control with this class of compounds. In addition, the C–X (X = Br, Cl) chain-end bonds of acrylates are typically stronger than those of methacrylates (Table 5). As a result, activation in the O-ATRP of acrylates is also more challenging. For these reasons, the polymerization of acrylates to form well-defined polymers (controlled molecular weights and low Ð) while achieving a high I* using this method was limited for several years.


Structures of acrylate monomers polymerized by O-ATRP.

Table 5.

Computed and Experimental Bond Dissociation Energies for Chain-end C─X (X = Br, Cl) Bonds of Polymers from Common Vinyl Monomers

monomerhalideΔG°a (kcal mol−1)BDEb (kcal mol−1)acrylonitrileBr47.2Cl55.8methyl methacrylateBr49.4Cl57.4styreneBr50.3Cl58.8methyl acrylateBr51.8Cl60.4vinyl ketoneBr53.3Cl61.3dimethyl acrylamideBr54.2Cl62.2vinyl chlorideBr55.965Cl65.979.5vinyl acetateBr59.5Cl69.4isobutyleneBr60.170Cl6984.1propyleneBr61.771.5Cl71.284.6vinyl etherBr63.2Cl72.1

Highlighting these challenges, many early attempts to polymerize acrylates exhibited moderate or poor polymerization control, as indicated by Ð around or above 1.5 and I* deviating significantly from 100%. Perhaps the first example of an acrylate polymerized by O-ATRP came in the seminal report by Miyake and Theriot, which used a PMMA macroinitiator (Mn = 72.9 kDa, Ð = 1.3) and n-butyl acrylate to synthesize a block copolymer (Mn = 219 kDa, Ð = 1.7). However, as evidenced by the large increase in Ð, the polymerization of this monomer was not well controlled. Similarly, Yilmaz and Yagci developed a method for the copolymerization of butyl acrylate (by O-ATRP) and ε-caprolactone (by ring opening polymerization) using a bifunctional initiator. Again, the copolymerization exhibited signs of poor control (Mn = 30.6 kDa, Ð = 1.7).

In work by Chen and co-workers, a core substituted phenothiazine catalyst was used to polymerize several monomers from sulfonyl halide initiators. Included in these monomers were methyl acrylate and n-butyl acrylate. In the case of methyl acrylate, poor polymerization control was observed, primarily through high dispersity (Ð > 1.5). However, the polymerization of this monomer did exhibit linear pseudo-first-order kinetics, linear molecular weight growth, and decreasing Ð throughout the reaction, suggesting some degree of polymerization control may have been present. Instead, the polymerization of n-butyl acrylate produced polymers with lower dispersity (Ð = 1.4), although with reduced initiator efficiency (I* = 82%). Similar results were obtained by Zhou and Lou, who were able to polymerize poly(ethylene glycol) acrylate with moderate control (Ð = 1.4, I* ~ 50%) using a water-soluble phenoxazine catalyst.

It was not until 2020 that O-ATRP was able to access a wide range of acrylates in a well-controlled fashion. With the development of dihydroacridine catalysts and enabled through the use of a continuous flow reactor, the polymerization of n-butyl acrylate was finally reported with good control (Mn = 2.4 to 45.7 kDa, Ð = 1.2 to 1.4, and I* ~ 100%). Further, the versatility of this polymerization system was demonstrated through the polymerization of methyl acrylate, ethyl acrylate, t-butyl acrylate, 2-ethylhexyl acrylate, and ethylene glycol methyl ether acrylate (Table 6). In each case, good or moderate polymerization control was observed, supporting the viability of this method to access a wide range of acrylate monomers.

Table 6.

Results from the Polymerization of Acrylates by Buss et al. Using Dihydroacridine PCs

monomerMn (kDa) Đ I* (%)methyl acrylate8.11.381ethyl acrylate7.81.2105t-butyl acrylate12.11.2882-ethylhexyl acrylate16.81.5104ethylene glycol methyl ether acrylate12.31.4117

Shortly after this report, McCarthy et al. explored the ability of dihydrophenazine catalysts to also access the polymerization of acrylates. Through their investigations, the authors discovered that tuning the polarity of the reaction solution could greatly impact the polymerization process. In particular, lowering the solvent polarity allowed access to the controlled polymerization of acrylates through modification of a number of important catalytic properties (see , ). Additionally, evidence was found for an in situ side reaction involving substitution of the catalyst by the radical initiator. This core-substitution was then performed intentionally through preirradiation of a catalyst and initiator solution, followed by addition of the monomer and additional initiator to begin the polymerization. The preirradiation step was intended to generate the substituted catalyst prior to polymerization, such that this side reaction would not consume initiator undesirably once the polymerization began. This approach ultimately yielded improved polymerization control (namely I* closer to 100%), presumably by eliminating this side reaction. In addition, this method was broadly applicable to a number of acrylate monomers, including methyl acrylate (Mn = 9.1 kDa, Ð = 1.2, I* = 83%), ethyl acrylate (Mn = 9.7 kDa, Ð = 1.2, I* = 97%), n-butyl acrylate (Mn = 7.7–17.5 kDa, Ð = 1.1–1.4, I* ~ 100% to 180%), t-butyl acrylate (Mn = 10.4 kDa, Ð = 1.2, I* = 115%), 2-ethylhexyl acrylate (Mn = 12.4 kDa, Ð = 1.2, I* = 115%), ethylene glycol methyl ether acrylate (Mn = 9.8 kDa, Ð = 1.4, I* = 127%), isobornyl acrylate (Mn = 10.3 kDa, Ð = 1.3, I* = 101%), and dicyclopentanyl acrylate (Mn = 15.4 kDa, Ð = 1.4, I* = 127%).

In work by the Wang group, acrylic acid was polymerized from the surfaces of nanoparticles using O-ATRP mediated by PhenS-Ph. While polymerization control was not evaluated in this report, the successful polymerization of this monomer was verified by FT-IR and transmission electron microscopy. In 2021, Qian, Han, Zhang, and co-workers demonstrated the polymerization of hexadecyl acrylate from functionalized cellulose-based fibers, although polymerization control was again not evaluated. Finally, an acrylate-based inimer similar to that reported for methacrylates was employed to synthesize hyperbranched polymers by O-ATRP. Gel permeation chromatography (GPC) analysis of the resulting polymer revealed broad molecular weight distributions throughout the polymerization, suggesting poor control over the polymer molecular weight and Ð. While a microemulsion polymerization was also attempted with this monomer to gain better control over the polymerization, it was ultimately unsuccessful as evidenced by the formation of unstable latexes and bimodal molecular weight distributions.

4.3. Acrylonitrile

Early in the development of O-ATRP, Matyjaszewski and co-workers reported on the use of various phenothiazine catalysts for the polymerization of acrylonitrile. While several new catalysts were developed for this application, namely phenothiazines with new N-aryl substituents, PhenS-Ph was ultimately the most successful. Using PhenS-Ph, the polymerization of acrylonitrile was moderately controlled, with the resulting polymer exhibiting Mn = 1.7–4.4 kDa and Ð = 1.4–1.9. Further, through the synthesis of a block copolymer with MMA, good chain-end fidelity was demonstrated with this system.

In 2017, Chen and Liu showed EY could also mediate the polymerization of this monomer using a benzenediazonium tetrafluoroborate initiator. In the homopolymerization of acrylonitrile, moderate polymerization control was again achieved, with the product poly(acrylonitrile) exhibiting Mn = 73–153 kDa, Ð = 1.2–1.6, and I* generally below 10%. In addition, these authors synthesized a series of statistical copolymers with acrylonitrile, using monomers such as MMA (5 mol %, Mn = 101 kDa, Ð = 1.3), methyl acrylate (15 mol %, Mn = 93 kDa, Ð = 1.3), n-butyl acrylate (5 mol %, Mn = 105 kDa, Ð = 1.4), styrene (5 mol %, Mn = 45 kDa, Ð = 1.2), and itaconic acid (5 mol %, Mn = 74 kDa, Ð = 1.2). In later work, this method was expanded to a number of other photocatalysts, such as rhodamine B, erythrosin B, and fluorescein. However, EY generally gave the best polymerization control and was ultimately chosen for subsequent experiments.

4.4. Acrylamides

In addition to acrylates and methacrylates, a handful of acrylamides have been polymerized by O-ATRP (Figure 22), although generally without much polymerization control. For example, Li and co-workers used surface-initiated O-ATRP to graft N-isopropylacrylamide (NIPAM) to the surface of SBA-15 nanoparticles. The resulting functionalized nanoparticles exhibited Mn = 13.4 kDa and Ð = 2.3, suggesting poor polymerization control with this monomer. When instead MMA was polymerized in the same system, polymers consistently showed Ð = 1.2–1.3, indicating a higher degree of polymerization control with this methacrylate monomer. Around the same time, Yilmaz and Yagci disclosed the concurrent O-ATRP and ring opening polymerization of NIPAM and ε-caprolactone, respectively, using a bifunctional initiator. However, the resulting polymer again exhibited signs of poor polymerization control (Mn = 33.2 kDa, Ð = 1.5). Slightly better polymerization control (Mn = 5.1–14.0 kDa, Ð = 1.4–1.5) was obtained by Hu and Wang in their attempts to synthesize block copolymers with NIPAM, although even this system gave Ð near the limit of control (Ð = 1.5).


Structures of acrylamide monomers polymerized by O-ATRP.

In addition to NIPAM, others have attempted the polymerization of acrylamide using O-ATRP. In an early example, researchers employed this monomer for the functionalization of Au electrodes for lead ion detection. Because the resulting polymer was surface-bound, it was not characterized to evaluate polymerization control. In a similar example, Sun and co-workers copolymerized acrylamide and N,N-methylene bisacrylamide, a bifunctional acrylamide cross-linker, to produce molecularly imprinted electrochemical sensors. Again, polymerization control was not evaluated due to the nature of the product polymer. More recently, Swisher et al. attempted the polymerization of N,N-dimethylacrylamide using newly developed phenoxazine PCs, although this polymerization showed little evidence of a controlled process. Instead, Liu and Yi employed a water-soluble benzophenone derivative for the homopolymerization of acrylamide in water, which gave poly(acrylamide) with Mn = 2.7–37.5 kDa, Ð = 1.4–1.5, and I* ~ 70% to 100%. Excitingly, these results suggest acrylamides have potential to be polymerized in a well-controlled fashion, assuming a suitable catalyst system can be developed for this monomer family.

4.5. Styrene and 4-Vinylpyridine

One common monomer that has largely remained elusive in O-ATRP is styrene, presumably because styrene can be a triplet quencher or because the dormant alkyl halide is more thermodynamically challenging to reduce (Figure 23). In 2014, this monomer was first accessed using perylene to synthesize a PMMA/poly(styrene) (PS) copolymer with moderate control (PMMA: Mn = 72.9 kDa, Ð = 1.3; PMMA-b-PS: Mn = 165 kDa, Ð = 1.4). Improved results were obtained by Yilmaz and Yagci, who reported the concurrent copolymerization of styrene and ε-caprolactone through O-ATRP and ring opening polymerization, respectively. In this system, the PC for O-ATRP was again perylene, and the resulting copolymer was obtained with Mn = 14.1 kDa and Ð = 1.2. However, many subsequent catalyst systems have been unable to polymerize styrene in a controlled fashion.


Structures of styrene monomers polymerized by O-ATRP and structure of 4-vinylpyridine.

In 2018, Kim, Gierschner, Kwon, and co-workers provided one example of styrene being polymerized by O-ATRP with moderate control (Mn = 8.7 kDa, Ð = 1.4, I* = 90%). However, it was not until 2019 in a report by Jessop and Cunningham that styrene was polymerized with the level of control expected for O-ATRP. In this work, the authors developed a new dihydrophenazine PC with pH sensitive functionalities for catalyst recycling. In addition to achieving this goal, they showed this catalyst could mediate the polymerization of styrene with good control (Mn ~ 18 kDa, Ð = 1.1–1.2) for the first time. Unfortunately, it remains unclear how the properties of this phenazine differ from those of other PCs and why this PC is successful in the polymerization of styrene when others are not. As such, further investigation of this catalyst system is warranted and could reveal important catalyst design principles for accessing new monomers in the future.

In addition to styrene, the analogous 4-vinylpyridine has also been employed in O-ATRP, although with relatively little emphasis on its controlled polymerization. Generally, this monomer has been employed for the production of self-healing hydrogels through grafting to a range of nanoparticle surfaces, and polymerization control has not been evaluated.- However, work by Nguyen, Truong, and co-workers showed that both PhenS-Py and pyrene can successfully polymerize 4-vinylpyridine to high conversions (up to 92%) with good polymerization control (Mn = 7.2–14.5 kDa, Ð = 1.1–1.2, I* = 95% to 102%). As such, the polymerization of 4-vinylpyridine by O-ATRP shows promise and warrants further investigation in the future.

Finally, 4-vinylbenzyl bromide, a styrene-based inimer, has also been polymerized by O-ATRP, this time in a copolymerization with styrene to synthesize styrene-based hyperbranched polymers. Unsurprisingly, when the product polymers were characterized by GPC, they exhibited high dispersity (Ð > 3), although the molecular weight of the polymers did show a dependence on the amount of inimer added to the reaction.

4.6. Vinyl Cyclopropanes

Another interesting monomer family that has received attention in O-ATRP has been vinyl cyclopropanes (Figure 24). Notably, these monomers contain coordinating functionalities, which may interact with metal catalysts in traditional ATRP and limit control in their polymerizations.- In O-ATRP, however, this issue is circumvented by the use of organic catalysts, which cannot coordinate with these monomers and are therefore better suited for these polymerizations. In 2019, the first application of O-ATRP to vinylcyclopropanes was reported by Chen et al. using phenoxazine and dihydrophenazine PCs. The dihydrophenazine PCs, PhenN-2N and PhenN-PhCF3, showed particularly good control in the polymerization of ethyl vinyl cyclopropane, producing polymer with Mn = 11.6–79.5 kDa, Ð = 1.1–1.4, I* = 91% to 127%. In addition, it was discovered that varying the reaction conditions could provide control over an intramolecular rearrangement of the polymer backbone, although this feature will be discussed further in a subsequent section (see , ).


Structures of vinyl cyclopropanes that have been polymerized by radical ring opening polymerization.

Through variation of the ester functionalities on the vinyl cyclopropane, tolerance for a wide range of functional groups was demonstrated, including alkyl chains, aromatic groups, and an alkyl chloride moiety. In general, most of the polymers synthesized showed excellent polymerization control, with Mn = ~ 20–50 kDa, Ð = 1.1–1.2, I* ~ 80% to 110%. In addition, later work expanded the scope of vinyl cyclopropanes to include new symmetric and asymmetric monomers, which featured functionalities ranging from cyano groups to natural products, and even poly(dimethylsiloxane) polymer chains. Through polymerization of the latter monomers, brush polymers were accessed with similar control (Mn = 67.4–309 kDa, Ð = 1.1–1.5, I* = 49% to 316%) relative to previous vinyl cyclopropane monomers.

4.7. Other Monomers

Finally, a handful of other monomers (Figure 25) have been investigated in O-ATRP, although generally as comonomers in conjunction with other monomers discussed in preceding sections. For example, when Niu and co-workers explored the polymerization of azide-containing methacrylates, they synthesized statistical copolymers with itaconic acid, ethyl vinyl ether, and butyl vinyl ether. In all three cases, polymerizations were performed with 20 mol% of the methacrylate and gave moderate polymerization control (Ð ~ 1.3–1.4), although the methacrylate content in the product polymer ranged from 25% (ethyl vinyl ether) to 67% (itaconic acid). In another example, Chen and co-workers synthesized statistical copolymers of acrylonitrile (95 mol %) and itaconic acid (5 mol %), which showed good polymerization control (Mn = 74 kDa, Ð = 1.2). However, only one example of this polymerization was reported, after which this system was not investigated further.


Structures of miscellaneous monomers polymerized by O-ATRP.

Vieira and co-workers explored the polymerization of d-limonene using benzophenone and thioxanthene-2-one. While the resulting polymers exhibited low dispersity (Ð = 1.1–1.2), monomer conversion was typically low (~6% to 12%) and seemed to plateau after several hours. As a result, polymer molecular weight was also quite low (Mn < 1 kDa), suggesting initiation had occurred but that the PCs could not reactivate the dormant polymer chains after deactivation. Given that the chainend of the polymer should be an unactivated alkyl-bromide, it is likely that reduction potential of the polymer chain-end is too low [E°(PnBr/PnBr•−) < −2.0 V vs SCE] to be reduced by the PCs in this work. Alternatively, it is also possible that the propagating polymer in these reactions underwent rapid termination, which would result in similar polymerization kinetics and observations of limited monomer conversion. However, further investigation is necessary to understand this system.

5. APPLICATIONS OF O-ATRP

One of the primary advantages of any controlled radical polymerization method is the ability to produce functional materials. In the following section, we outline the various ways in which O-ATRP has been employed in this respect, ranging from the synthesis of various polymer architectures to the surface functionalization of nanoparticles and electrodes.

5.1. Synthesis of Block Copolymers

The synthesis of block copolymers by O-ATRP has been achieved using several strategies. Perhaps the simplest strategy involves the chain-extension of polymers also produced using this method, which takes advantage of their bromide chain-end functionality to initiate further O-ATRP reactions (Figure 26a). This strategy can be implemented in one of two ways. In the first, monomers can be added sequentially to a polymerization, such that each block of the copolymer is formed one at a time in the same pot. While operationally simple, this method requires high conversion of the first block to achieve a well-defined transition from one block to the other. However, it is also common to lose some polymerization control at high monomer conversions, a feature that is commonly seen in ATRP methods, which can complicate the synthesis of well-defined copolymers by this method. As such, an alternative approach involves the synthesis and isolation of the first block, often called a macroinitiator, followed by chain-extension of the macroinitiator in a separate polymerization to generate the desired copolymer.


(a) Broad depiction of block copolymer synthesis through sequential monomer (M) addition using O-ATRP. (b) The impact of end-group fidelity on Ð of the block copolymer. (c) Structures of block copolymers synthesized using perylene by Miyake and Theriot in 2014.

Because the formation of the second polymer block is dependent upon the presence of bromine chain-end groups in the first block, this method is often used to evaluate chain-end group fidelity in a given polymerization. The principle behind this experiment is depicted in Figure 26b. For an ideal polymer sample in which all the chain-end groups are retained, it can be expected that chain-extension by a well-controlled polymerization will result in complete conversion of the macroinitiator to the desired block copolymer. In this idealized case, analysis of the GPC trace of the copolymer should reveal a narrow, monomodal peak, indicating complete chain-extension. If, instead, some portion of the first polymer block is unfunctionalized, chain-extension will primarily result in the formation of two polymer species within the sample: the unfunctionalized first polymer block and the chain-extended copolymer. The presence of this unfunctionalized polymer can sometimes be observed by GPC and is indicated by the observation of a bimodal peak in the chromatogram of the chain-extended polymer.

Of course, other reasons may also exist for the observation of these features during GPC analysis. For example, a polymer chromatogram could be multimodal simply due to poor polymerization control. In addition, it is not uncommon for a copolymer to exhibit different hydrodynamic properties than the corresponding macroinitiator, which could lead to complications during GPC analysis. For this reason, Junkers and Michels have recommended against the use of these chromatograms alone as evidence for successful chain-extension, as they can sometimes be misleading. Instead, other methods can provide more reliable evidence, such as multi-angle light scattering in which absolute polymer molecular weight can be determined without interference from the polymer architecture or structure.

For examples of this block copolymer synthetic strategy being applied in O-ATRP, a number of literature reports exist. The first examples are once again found in the seminal reports by Miyake and Theriot and Hawker, who made copolymers from PMMA macroinitiators and a number of other comonomers. In the former report, butyl methacrylate, butyl acrylate, and styrene were used to form the second polymer block (Figure 26c). However, GPC analysis revealed a significant portion of unreacted PMMA, suggesting poor chain-end fidelity in O-ATRP mediated by perylene. Instead, the report by Hawker showed high conversion of the PMMA macroinitiator in the copolymerization with benzyl methacrylate, suggesting better chain-end fidelity in the presence of PhenS-Ph. This conclusion was further supported with copolymerizations mediated by photoATRP and Cu catalyzed ATRP, which capitalized on the complementary strengths of these methods to form copolymers with methyl acrylate and styrene, respectively.

By expanding on this work, subsequent reports showed a number of methacrylate monomers could be used to form copolymers with PMMA., In addition, copolymers can be synthesized with other macroinitiators, such as poly-(acrylonitrile), poly(n-butyl acrylate), and poly(N-isopropylacrylamide). In the latter case, a copolymer of NIPAM and t-butyl acrylate was synthesized, after which the acrylate block was hydrolyzed to acrylic acid to achieve a copolymer that would have otherwise been challenging to synthesize.

Finally, copolymers with more than two blocks have also been achieved through repetitive polymer isolation followed by chain-extension. For example, Cole et al. synthesized a triblock methacrylate copolymer using aryl core-substituted dihydrophenazine PCs. Similarly, Buss et al. demonstrated the synthesis of a triblock acrylate copolymer using dihydroacridine PCs, highlighting the excellent polymerization control obtained in this method.

While a number of interesting copolymers can be obtained using the strategy discussed above, it is inherently limited to the incorporation of monomers accessible through O-ATRP. For this reason, another common strategy involves the postpolymerization modification of polymers obtained by other methods, such that they can be used as macroinitiators in O-ATRP (Figure 27a). For an example of this strategy, one can look to the work by Son and co-workers reported in 2018, in which copolymers of poly(ethylene glycol) (PEG) were synthesized by addition of a bromoisobutyrate group to the PEG chain-end followed by O-ATRP of glycidyl methacrylate (Figure 27b). While polymerization control in this system was generally poor (Ð > 1.5), later work by the same group showed phenoxazine PCs could yield similar copolymers with greater control (Ð < 1.5). In a similar approach, Nguyen and co-workers synthesized copolymers of poly(3-hexylthiophene) and various methacrylates with good polymerization control (Figure 28).


(a) General scheme of block copolymer synthesis by functionalization of a polymer and chain-extension using O-ATRP. (b) Synthesis of amphiphilic block copolymers by functionalization and chain extension of poly(ethylene glycol).


Synthesis of P3HT-b-PMMA using O-ATRP by chain extension of a P3HT macroinitiator.

In another popular strategy, multifunctional initiators are employed to perform orthogonal polymerizations of different monomers in one pot (Figure 29a), allowing access to a range of monomers inaccessible to a single polymerization method. These orthogonal polymerizations can either be performed separately or concurrently, depending on the desired conditions and compatibility of the chosen synthetic methods. For example, Theriot et al. showed dihydrophenazines could mediate PET-RAFT followed by O-ATRP to synthesize poly(acrylate-block-methacrylate) copolymers (Figure 29b).


(a) Synthesis of block copolymers through two orthogonal methods in one pot. (b) Sequential PET-RAFT and O-ATRP to synthesize PMA-b-PMMA.

Instead, Yilmaz and Yagci used a bifunctional initiator similar to one reported by the Boyer group comprised of an alcohol and a bromoisobutyrate moiety to perform concurrent ring opening polymerization and O-ATRP (Figure 30). Using this approach, ε-caprolactone was copolymerized with a series of vinyl monomers, including MMA, n-butyl acrylate, styrene, and NIPAM. A triblock copolymer was also synthesized through chain-extension of the ester block using l-lactide, demonstrating the versatility of this method and its ability to produce highly tunable copolymers. Shortly thereafter, Yilmaz extended this work to a trifunctional initiator, one bromoisobutyrate group tethered to two alcohols, allowing for the synthesis of star polymers through this same approach.


Concurrent O-ATRP and ROP to synthesize PMMA-b-PCL.

In work published by the Hawker and Read de Alaniz groups in 2018, several methods were employed to synthesize highly functionalized copolymers (Figure 31). First, a series of monomers and initiators suited for ATRP were synthesized, bearing furan-protected maleimides. These compounds were then employed in O-ATRP to synthesize methacrylate polymers with maleimide end-groups, as well as copolymers with maleimide pendant groups. In the latter case, the versatility of O-ATRP was shown in the synthesis of a tetrablock copolymer, where polymerization of the maleimide functionalized monomers was enabled by the mild reaction conditions found in O-ATRP. Finally, polymers bearing furan protected maleimide end-groups were further modified using Diels–Alder chemistry to install a PEG block within the copolymer.


Synthesis of initiators and block copolymers functionalized with furan-protected maleimides.

5.2. Synthesis of Graft Polymers

Regarding the synthesis of polymers with higher-order architectures, O-ATRP has primarily been used for the synthesis of block copolymers. However, several examples exist of O-ATRP being used to produce more complex polymer architectures, such as graft polymers. In O-ATRP, this architecture is often achieved using a grafting-from approach, where a polymer backbone is functionalized with an alkyl halide group from which O-ATRP can be initiated. An excellent example of this approach was reported in 2018 by Chen and co-workers. In this work, the authors first developed aryl sulfonyl halides as an initiating system for O-ATRP. Once the success of this method was demonstrated using small molecule initiators, it was recognized that poly(styrene) could be functionalized using this approach to yield sulfonyl halide initiating sites on the pendant phenyl groups of the polymer. By then performing O-ATRP from these sites, poly(styrene-graft-acrylate) polymers could be prepared through a grafting-from approach.

In another example, the chloride functionality in the backbone of poly(vinylidene fluoride-co-chlorotrifluoroethylene) [P-(VDF-co-CTFE)] was exploited for the synthesis of graft copolymers for electronic applications (Figure 32). Using PhenS-Ph as the catalyst, monomers such as MMA, methyl acrylate, and n-butyl acrylate were grafted from P(VDF-co-CTFE). NMR and GPC analysis demonstrated the success of the grafting process, with graft contents ranging from 5% up to 38%. However, these polymerizations showed poor control (Ð » 1.5), possibly due to challenges associated with the use of Cl− in O-ATRP (see , ). In support of this hypothesis, later work showed P(VDF-co-CTFE) could be dechlorinated with high yield under these conditions, suggesting activation of the polymer C─Cl bond is feasible. In addition, Hu, Fang, Lu, and co-workers showed this polymerization could also be mediated by p-anisaldehyde, offering an inexpensive catalyst system for the synthesis of these materials.


O-ATRP initiated from the backbone of P(VDF-co-CTFE).

Expanding on the polymeric backbones available for graft copolymer synthesis, Wang and Chu showed ethyl cellulose could be modified by installation of bromophenylacetate functionalities for use as an O-ATRP macroinitiator. After grafting various methacrylates from this polymer backbone (Figure 33), a series of graft copolymers with predictable molecular weights but poor dispersities (Ð ~ 1.7) was obtained. In later work, it was shown these polymers could be further modified to produce cellulose-based thermoset elastomers. Capitalizing on the ability to incorporate furfuryl methacrylate into the polymeric arms of the graft copolymer, the authors used Diels–Alder chemistry to create dynamic cross-links within the polymer network. The resulting materials exhibited shape recovery as well as self-healing properties due to the dynamic nature of the cross-linked network.


Synthesis of graft polymers from cellulosic materials using O-ATRP.

In addition to the grafting-from approach described above, grafting-through has also been used in the synthesis of graft polymers by O-ATRP. In this method, macromonomers, polymers bearing polymerizable end-groups, are synthesized and then polymerized to create the desired graft polymer. As such, the graft polymer backbone is formed as the macromonomers are linked together in the polymerization. Perhaps the first example of this strategy in O-ATRP was reported by Matyjaszewski and co-workers, who synthesized poly(ethylene oxide) methacrylate lithium sulfonyl(trifluoromethylsulfonyl)-imide (PEOMA-TFSI−Li+), a PEG-based zwitterionic monomer for battery applications (Figure 34). Using O-ATRP, this monomer was polymerized with good control (Ð = 1.2–1.4) to yield polymers with Mn tunable from 1 to 30 kDa. Through electrochemical testing, these polymers were shown to exhibit good conductivity, good electrochemical stability, and potential to suppress dendrite growth in batteries.


Synthesis of polymeric electrolytes using O-ATRP.

A similar approach to graft copolymer synthesis was later reported by Chen et al., who synthesized poly-(dimethylsiloxane) functionalized vinylcyclopropanes for radical ring-opening polymerization. Upon polymerization of these monomers, polymers could be obtained with Ð < 1.5 and with a range of molecular weights (Mn = 67.4–309 kDa). Further, through control of the polymerization conditions, intramolecular reorganization of the polymer backbone could be controlled to obtain a primarily linear or cyclized structure. While this feature will be discussed further in a subsequent section (see , ), it is worth nothing here that control over this backbone structure could enable future investigations into the impact of this structural feature on the graft polymer properties.

5.3. Synthesis of Hyperbranched Polymers

In 2017, Yagci and co-workers showed hyperbranched polymers could be synthesized via O-ATRP by the simultaneous copolymerization of MMA and a methacrylate-based inimer (see , ). In this work, the inimer content was varied from 9% to 27%, with the resulting polymers exhibiting Mn ranging from 101 to 604 kDa and Ð = 2.7–6.2. In addition, the resulting polymers could be chain-extended with styrene (Figure 35), suggesting good chain-end fidelity despite these metrics of poor polymerization control. To improve upon this system, Gao and co-workers developed a similar polymerization method in microemulsion, hypothesizing that the spatial constraints created by the microemulsion would yield lower Ð branched polymers. Indeed, when these polymerizations were attempted under these constraints, the resulting hyperbranched polymers exhibited much narrower molecular weight distributions (Ð ~ 1.7–2.2), supporting the authors’ hypothesis. Importantly, this work also represents the only example of O-ATRP performed in microemulsion, creating opportunities for further development in this area.


Synthesis of hyperbranched polymers by O-ATRP through the copolymerization of MMA and a bifunctional methacrylate monomer.

5.4. Synthesis of Star Polymers

First explored by Buss et al. in 2018, the synthesis of star polymers by O-ATRP has received attention from several groups in recent years. While several approaches to star-polymer synthesis exist, methods employing O-ATRP have primarily focused on the core-first approach (Figure 36a). In essence, this strategy involves the use of a multifunctional initiator, which during the polymerization initiates the growth of several polymer arms tethered together at the initiator center. Because the number of initiating sites on the initiator can be precisely tuned, the number of polymer arms within the star architecture can also be controlled exactly.


(a) Star polymer synthesis via O-ATRP using multifunctional initiators. (b) O-ATRP from a multiarmed bromide-containing initiator. (c) Concurrent O-ATRP and ROP to produce three-armed copolymers. (d) O-ATRP from functionalized cellulosic materials to produce star polymers.

Highlighting the versatility of this strategy, Buss et al. synthesized star polymers with 2–8 arms using bromoisobutyrate-based multifunctional initiators (Figure 36b). The primary monomer chosen for these investigations was MMA, given the previous success of this monomer in O-ATRP. However, through chain-extension of PMMA star polymers, benzyl methacrylate was also incorporated into these materials. In general, polymerization control was quite good given the complexity of this architecture, with Mn = 18.3–68.4 kDa, Ð = 1.2–1.9, and I* ~ 100%. As the number of polymer arms increased, polymerization control was usually lost, with Ð and I* increasing undesirably. However, this observation is not surprising and can be expected for this type of polymer architecture. In later work, similar results were obtained using aryl core-substituted dihydrophenazine PCs, although at significantly reduced catalyst loadings relative to the original report above.

In a similar approach, Yilmaz developed star-shaped polymers of MMA or styrene with ε-caprolactone (Figure 36c). In this case, the initiator had to be modified to include alcohols, which enabled the ring opening polymerization of the ester-based monomer. Nonetheless, the resulting polymers were produced with moderate control (Mn ~ 10–20 kDa, Ð ~ 1.3–1.4), highlighting the ability of O-ATRP to operate effectively in the presence of other polymerization systems.

In 2020 Pang and Qiao reported on the functionalization of β-cyclodextrin with bromophenylacetate moieties to form the core of amphiphilic star shaped polymers (Figure 36d). Using O-ATRP, t-butyl acrylate was polymerized from this core, followed by MMA to form diblock copolymer arms. The poly(t-butyl acrylate) block was then converted to poly(acrylic acid) by hydrolysis, creating star polymers with hydrophilic cores and hydrophobic shells. In every case, the polymers produced showed high levels of polymerization control, including predictable molecular weights and Ð typically below 1.2.

5.5. Surface-initiated O-ATRP

The first example of surface-initiated O-ATRP (SI–O-ATRP) came relatively early in the development of O-ATRP. Reported by Read de Alaniz and Hawker in 2016, this method was first developed for the light-mediated growth of polymers tethered to silicon surfaces (Figure 37a,​b). Although this method has since been extended to a number of other materials and surfaces, such as nanoparticles (Figure 37c), the principle generally remains the same: an alkyl halide initiator is tethered to a surface, after which O-ATRP is performed to generate polymer functionalities at that surface. In the report by Read de Alaniz and Hawker, Si wafers were functionalized with bromoisobuty-rate groups, which allowed for the surface-initiated polymerization of MMA (Figure 38a). Unsurprisingly, this process was shown to be dependent on irradiation, where increasing the intensity of the light source increased the rate of polymer brush growth over time. In addition, the chain-end fidelity of this method was demonstrated through the synthesis of block copolymers, the success of which was determined using X-ray photoelectron spectroscopy (XPS) to identify fluoride functionalities in the second polymer block.


(a) General approach to surface functionalization using O-ATRP. (b) The use of photomasks to produce patterned surfaces. (c) Functionalization of nanoparticles using SI–O-ATRP.


Two approaches to SI-O-ATRP from silicon for the polymerization of MMA (a) and MAA (b).

Given the dependence of this polymerization on irradiation, the authors proposed surface patterning could be achieved by employing photomasks to control which parts of the surface were irradiated (Figure 37b). Indeed, when polymerizations were performed in the presence of a photomask, precise patterns with features on the micrometer scale could be achieved with excellent reliability, supporting the feasibility of this approach. In fact, this strategy could also be applied to the synthesis of block copolymers, allowing hierarchical patterns to be produced.

Expanding on this method, subsequent research focused in part on the incorporation of new monomer functionalities. For example, work by Junkers and co-workers showed methacrylic acid could be grafted to Si surfaces (Figure 38b), the presence of which was probed using XPS and secondary ion mass spectrometry (SIMS). In another case, Hawker and co-workers developed methacrylates based on emissive Ir complexes, which were polymerized through surface initiated O-ATRP to produce patterned, emissive surface coatings. This second example also showed these polymerizations could be performed under ambient conditions (i.e., under air), as long as a glass coverslip was placed above the reaction solution to minimize the diffusion of air into the polymerization.

Of course, Si is not the only material of interest for surface functionalization, so some of the research in this field has focused on expanding surface-initiated O-ATRP to other materials. In one example, Tang, Xu, Zhou, and co-workers developed the surface-initiated polymerization of semifluorinated methacrylates on indium tin oxide (ITO) and fluorine doped tin oxide (FTO) glass to improve the durability and hydrophobicity of these materials. In a similar approach to that reported by Hawker, the surfaces of ITO and FTO glass were functionalized with a bromoisobutyrate initiator fragment, from which O-ATRP was performed with PhenS-Ph as the catalyst. Through optical characterization of the functionalized surfaces, it was shown the polymer coating had minimal influences on the transmittance of the glass. Further, when the conductivity of the polymer functionalized ITO glass was characterized, it was discovered that films up to 1 μm in thickness had a minimal impact on the conductivity of the ITO. As such, the authors demonstrated successful functionalization of these materials with minimal impact on their desirable properties.

In addition, surface-initiated O-ATRP has been used to tether polymer films to a number of electrode surfaces for electrochemical sensing applications. The first report of this type came in 2017 from the Sun group, who prepared poly(acrylamide-block-methacrylic acid) copolymers tethered to gold electrodes for the detection of lead ions. Like previous reports, the presence of the surface-bound polymer was probed using XPS. In addition, testing of the sensor revealed it could operate over a large linear range ([Pb2+] = 10−11–10−4 M) with a low limit of detection (2.5 × 10−12 M) and excellent selectivity for Pb2+ in the presence of other metal ions.

Unfortunately, not all sensors exhibit high selectivity for the target analyte, and creating a selective sensor can sometimes be challenging. To address this issue, one strategy uses molecularly imprinted polymers, where polymerizations are performed in the presence molecular template to create a polymer network around the template. After the polymerization, the template is removed, leaving behind a cavity in the polymer network designed to selectively bind the template molecule during sensing applications (Figure 39a).


(a) General design strategy for selective sensors by electrode functionalization using O-ATRP and molecular imprinting. (b) Copolymerization of acrylamide and a bifunctional acrylamide monomer using erythromycin as the molecular imprinting template. (c) Copolymerization of MAA and a bifunctional methacrylate using histamine as the molecular imprinting template.

The first application of O-ATRP in molecular imprinting was reported by Sun and co-workers, who used fluorescein to copolymerize acrylamide and an acrylamide-based cross-linker on modified Au electrodes (Figure 39b). Erythromycin was chosen as the molecular template, as this molecule was also the target analyte the authors ultimately wanted to measure. After the polymerization and removal of the template molecule, the selectivity of the functionalized electrode was tested by measurement of a range of analytes. Interestingly, the greatest response was obtained for erythromycin, demonstrating the success of this method. In addition, the reported sensor featured a large linear range ([erythromycin] = 10−8–10−1 M) and low limit of detection (3.2 × 10−9), suggesting the sensor is both selective and sensitive.

In a similar approach, Junkers and co-workers prepared molecularly imprinted polymers for the detection of histamine using O-ATRP catalyzed by PhenS-Ph. This time, methacrylic acid was copolymerized with a methacrylate-based cross-linker from a modified titanium electrode (Figure 39c). When the sensitivity of the resulting electrode was compared to that of an unfunctionalized electrode, it was shown that molecular imprinting significantly increased the sensitivity of the electrode to histamine. However, when the selectivity of the sensor was investigated, it was discovered that histidine could also produce an interfering response, potentially limiting the reliability of this sensor.

In addition to these reports, several other examples exist of O-ATRP being applied to generate polymeric coatings tethered to electrode surfaces. For instance, Kong and co-workers used surface initiated O-ATRP to grow ferrocenylmethyl methacrylate polymers from surface-bound DNA on Au electrodes. Instead, Chen, Bain, and co-workers used surface initiated O-ATRP catalyzed by eosin Y to polymerize glycidyl methacrylate at the surface of carbon nanotubes, which were then used as nanoprobes to improve the detection of carcinoembryonic antigen and α-ferroprotein through an electrochemical method. Finally, in an example from the Sun group, immunoglobulin G imprinted polymers were prepared at the surface of a modified Au electrode using O-ATRP catalyzed by fluorescein. In each of these examples, the wide range of functionalities tolerated both within monomers and other components of the sensors highlights the incredible versatility of O-ATRP.

Another common application of surface-initiated O-ATRP is in the functionalization of nanoparticles. Here, the use of a controlled polymerization is critical, as termination reactions such as radical coupling can rapidly lead to discrete nanoparticles coupling to each other and forming an interconnected network. As such, the high degree of polymerization control obtainable through O-ATRP positions this method well for use in the functionalization of discrete nanoparticles.

The first example of O-ATRP being applied in this fashion came again from Read de Alaniz and Hawker. In addition to functionalizing Si surfaces, these authors showed their method could be applied to SiO2 nanoparticles, the success of which was evaluated by transition electron microscopy (TEM). Following up on this work, which primarily used a bromoisobutyrate-derived initiator, Matyjaszewski showed similar results could be obtained by using a bromophenylacetate-based initiator (Figure 40). More importantly, Matyjaszewski’s report showed that greater polymerization rates and grafting densities could be obtained using bromophenylacetate initiators, enabling the density of polymer chains on the nanoparticle surface to be tuned.


SI-O-ATRP from silica nanoparticles using different initiators.

In an application of this method, Li and Wang demonstrated that SiO2 nanoparticles functionalized using O-ATRP could be employed in drug delivery, where a drug is encapsulated in the polymer network and then released through exposure to a stimulus (see , ). In this case, a pH sensitive polymer, poly(diethylamino ethyl methacrylate), was grafted to the SiO2 surface and loaded with quercetin, a potential anticancer drug. When the functionalized nanoparticles were then placed in solution, drug release was controlled by increasing the pH of the solution. Through protonation of the polymer pendant groups, electrostatic repulsions caused the polymer chains to expand, leading to the release of the encapsulated drug. Similar results were obtained when SBA-15 was used as the substrate for polymer grafting.

In related research, other groups have applied surface-initiated O-ATRP to the functionalization of other nanoparticle materials, such as hollow SiO2 spheres and mesoporous SiO2. In the former example, MMA and NIPAM were polymerized from the surface of SiO2 hollow spheres to improve their dispersibility in water. In the latter case, several groups have reported on the functionalization of SBA-15,,- a mesoporous SiO2 material developed by researchers at the University of California at Santa Barbara.

The first example using this material came from Li and co-workers in 2017, who showed SBA-15 functionalized with bromoisobutyrate moieties could be used to initiate the polymerization of MMA, dimethylamino ethyl methacrylate, and NIPAM. In each case, PhenS-Ph was employed as the O-ATRP catalyst. The resulting materials showed improved absorption of toluene for purification of contaminated aqueous solutions. In later work, Zhang and co-workers showed similar materials could also be obtained using fluorescein as the catalyst (Figure 41).


SI-O-ATRP from SBA-15 nanoparticles.

To understand how various reaction components impact these polymerizations, Zhang and co-workers performed a systematic investigation of the polymerization of MMA from SBA-15. To analyze polymerization control, the surface-bound PMMA was cleaved from the SBA-15 and characterized. As a result, the authors discovered the polymer Ð could be lowered through increasing the solvent quantity in the polymerization as well as increasing the amount of sacrificial amine used in conjunction with fluorescein. Interestingly, changing the quantity of added amine did not impact the molecular weight of the polymer. However, a link between the concentration of monomer and molecular weight was established, enabling predictable control over polymer molecular weight in this method.

Work by Xu, Zhang, Wei, and co-workers focused on the applications of these materials, showing that they can be used for drug delivery and biological imaging when functionalized with fluorescent groups. In this case, the fluorescent moiety was a surface-bound PhenS-Ph derivative, which also served as the PC for the copolymerization of PEG methacrylate and itaconic acid. When the polymerizations were complete, cell uptake was probed using optical and fluorescence microscopy, and pH-dependent drug release was demonstrated under acidic conditions.

In addition to these SiO2-based materials, a number of other nanoparticles have been modified using surface-initiated O-ATRP. These include ceria nanoparticles, nanodiamond, rare earth doped upconversion nanoparticles, magnetic Fe3O4 nanoparticles,, ferroelectric BaTiO3 nanoparticles, cellulose nanocrystals, and Eu3+ doped luminescent hydroxyapatite.- In particular, the functionalization of hydroxyapatite nanorods by O-ATRP has received considerable attention, with target applications including biological imaging and drug delivery. Often, surface functionalization of the nanorods is used to improve their solubility and cell-uptake properties, with the most common monomer being PEG methacrylate (Figure 42)., However, other monomers have also been employed, including 2-methacryloyloxyethyl phosphorylcholine, a zwitterionic methacrylate, and itaconic acid, which were used to transport cisplatin into HeLa cells.


Surface functionalization of hydroxyapatite nanoparticles for biological imaging via SI-O-ATRP.

One final application of surface-initiated O-ATRP that has been under development is in the synthesis of self-healing hydrogels. Typically, these materials have been achieved through the surface-initiated polymerization of 4-vinylpyridine on various nanoparticles, followed by the free radical polymerization of acrylic acid in the presence of these functionalized nanoparticles.- As a result, the pyridine moiety undergoes protonation by acrylic acid, generating ionomers that are electrostatically attracted to each other. Upon application of a force to the bulk material, these electrostatic interactions can be disrupted, leading to separation of the polymer chains and ultimately mechanical failure of the material. However, because these noncovalent interactions can be easily reestablished, self-healing properties are often observed (Figure 43).


(a) General principle behind self-healing materials produced by O-ATRP. (b) Polymerization of 4-vinylpyridine on silica nano-particles to produce self-healing hydrogels using O-ATRP.

In the first example of O-ATRP being used in this manner, SiO2 nanoparticles were functionalized with poly(4-vinyl-pyridine) using rhodamine B as the catalyst (Figure 43b). Since then, various other materials have also been employed, including cellulose nanocrystals, porous carbon nanospheres, and carbon nanotubes. However, regardless of the specific nanomaterial chosen, similar results are generally observed. The hydrogels produced can be broken and healed within a few hours, with up to 90% retention of the material’s original tensile strength after healing.

5.6. Synthesis of Polymers in Continuous Flow

While there are several advantages to using O-ATRP over other methods, it also features several limitations. Perhaps one of its biggest limitations is one that is experienced often in photochemistry: scalability. The issue with scaling photochemical reactions is that it is often very difficult to maintain uniform irradiation of the reaction solution at large reaction volumes (i.e., more than a few mL). To address this issue, many have employed photochemical flow reactors,,, where the reaction is passed through clear tubing surrounding a light source to ensure each part of the reaction is consistently and uniformly irradiated (Figure 44a). As a result, reactions performed in flow can in theory be scaled limitlessly, as long as the necessary reagents and reaction components can be supplied to the reactor.


(a) Synthesis of polymers by O-ATRP in a continuous flow reactor. (b) O-ATRP of methacrylates in flow. (c) O-ATRP of methacrylates in flow using PhenS-Ph as the PC. (d) O-ATRP of MAA in flow. (e) O-ATRP of acrylate monomers in flow.

As such, the issue of scalability in O-ATRP has been addressed to some degree using continuous flow reactors. The first report of this type came in 2017 from Ramsey et al. (Figure 44b), who showed various methacrylates could be polymerized in the presence of phenoxazines, phenazines, and even perylene in flow. In many cases, good polymerization control was demonstrated (Ð ≤ 1.2, I* ~ 100%) even at reduced catalyst loadings (1000 ppm: Mn = 6.1 kDa, Ð = 1.2, I* = 95%; 100 ppm: Mn = 7.1 kDa, Ð = 1.3, I* = 90%, both with PhenO-1N-BiPh). This reduction in catalyst loading was hypothesized to be possible due to the improved irradiation conditions in a continuous flow reactor, demonstrating another advantage to this reactor design. Further, molecular weight control was demonstrated by varying the residence time of the reaction, the amount of time it takes to pass through the full length of the reactor tubing, enabling the same control over polymer structure that can be obtained under batch reaction conditions by varying reaction time.

In subsequent work by Hu, Zhu, and co-workers, the application of O-ATRP using phenothiazine derivatives to a flow reactor was also investigated (Figure 44c). In this case, the authors investigated the effect of varying the reactor tubing diameter, as this factor could influence mixing and irradiation of the reaction solution. Ultimately, they found that the tubing diameter can significantly impact both Ð and I*, but that either too small or too large of a diameter can negatively impact polymerization control. In other words, an intermediate size exists where the best polymerization results can be achieved. In this case, a 2 mm inner diameter was optimal. In addition, the authors showed this method could be performed with a series of phenothiazine PCs, but that PhenS-Ph ultimately gave the best polymerization control in the O-ATRP of MMA.

Expanding on the scope of O-ATRP in flow, Rolando and co-workers showed eosin Y could also be employed with this reactor design. In the polymerization of MMA, this system gave moderate polymerization control, with good I* (I* ~ 100%) and moderate dispersity (Ð ~ 1.4). In addition, others have shown a wide range of monomers can be polymerized using O-ATRP in flow, such as methacrylates, methacrylic acid (Figure 44d), acrylates (Figure 44e), and styrene.

5.7. Metal Sensitive Applications of O-ATRP

One of the most commonly cited metal sensitive applications of ATRP is in electronics, which is also one of the areas where the superiority of O-ATRP over traditional, metal-catalyzed ATRP has been demonstrated. In the work by Zhang and co-workers, O-ATRP was employed to synthesize graft copolymers from the backbone of P(VDF-co-CTFE) (see , ). The ultimate goal of this work was to produce more efficient dielectric materials for energy storage applications, which the authors hoped to achieve by grafting insulating polymer chains to P(VDF-co-CTFE). In turn, it was hypothesized these insulating chains would reduce interactions between neighboring P(VDF-co-CTFE) domains, leading to less energy loss and improved electronic properties. However, the authors also hypothesized the use of metal catalyzed ATRP would be detrimental to the electronic properties of the product polymers, as residual metal ions remaining in the polymer matrix could migrate under an applied electric field and lead to undesirable energy loss. As such, they proposed O-ATRP would produce materials with enhanced performance relative to Cu catalyzed ATRP.

To test this hypothesis, Zhang and co-workers used both O-ATRP and Cu catalyzed ATRP to grow PMMA from the C─Cl bonds in the P(VDF-co-CTFE) backbone. The dielectric properties of the resulting materials were then tested, which revealed undesirable ion migration in the polymers prepared by Cu catalyzed ATRP. By contrast, this issue was not observed in the polymers prepared by O-ATRP, indicating superior electronic properties as hypothesized. As such, this report represents one of the first examples demonstrating a clear advantage of O-ATRP over traditional, metal catalyzed ATRP.

In another example, Matyjaszewski and co-workers employed O-ATRP for the synthesis of graft copolymer-based electrolytes for battery applications. While it is possible that metal contamination arising from traditional ATRP methods could also be problematic in this application, this issue was not evaluated in this work. Similarly, another possible application of O-ATRP in a metal-sensitive system has been toward functionalizing electrode surfaces for electrochemical sensors (ex., for the detection of Pb2+), although it has not be evaluated whether O-ATRP has a clear advantage over metal catalyzed ATRP. Moving forward in the development of this method, further studies in these areas directly comparing materials produced by metal catalyzed ATRP and O-ATRP could be beneficial.

Another commonly cited metal-sensitive application of ATRP is in biological materials, such as those for biological imaging and drug delivery. Indeed, a number of reports exist for the application of O-ATRP in these areas, ,,,,- with one notable example being that by Deng, Zhang, and Wei in 2017 (Figure 45). In this work, Eu3+ doped hydroxyapatite was functionalized using surface-initiated O-ATRP, such that the functionalized nanoparticles could be loaded with cisplatin and introduced into living cells. After HeLa cells were incubated with these materials, the fluorescent properties of the nanoparticles were exploited for cell imaging to confirm cell uptake. Further, the pH responsiveness of the polymer functionalities was employed to release cisplatin into the cell, demonstrating the potential for these nanoparticles to be used in drug delivery applications.


(a) One polymeric drug delivery strategy. (b) Functionalization of hydroxyapatite for drug delivery and biological imaging.

Although concerns of metal contamination in biological applications are often cited as a motivation for O-ATRP, it is worth noting this is a somewhat nuanced issue. For example, while some metals may pose toxicity concerns, not all metals are toxic, and this point certainly applies to the metal catalysts employed in ATRP as well. In fact, some have shown that ATRP can even be performed using nontoxic Cu dietary supplements. It is also worth noting that some of the catalyst families employed in O-ATRP (ex. phenazines and phenothiazines) are known to be biologically active,- although their effect in humans is still unclear. As such, further research is necessary to truly understand whether many of the catalysts employed in O-ATRP are biocompatible.

Another important area in which O-ATRP has a clear advantage over metal catalyzed ATRP is in the polymerization of coordinating monomers. In traditional ATRP, such monomers have typically been challenging because their coordination to the metal catalyst can alter catalytically relevant properties., As a result, the complex formed upon coordination may not be well suited to mediate ATRP. One way to address this issue is through the use of organic catalysts, such as in O-ATRP. The first example demonstrating this advantage was the seminal report by Hawker and co-workers, which showed PhenS-Ph could control the polymerization of dimethylamino ethyl methacrylate while a metal-based catalyst could not.

In addition to this example, Chen et al. showed various vinylcyclopropanes could be successfully polymerized by O-ATRP with excellent control over the resulting polymer structure., While this monomer family was previously polymerized using Cu catalyzed ATRP, the polymerizations were limited to low monomer conversion. Further, control over the backbone configuration, which can either be linear or undergo rearrangement to a cyclic structure (Figure 46), was poor in this method. In contrast, the polymerization of these monomers under O-ATRP conditions (Figure 46) yielded high monomer conversion (>90%) and excellent control over the backbone configuration (13% to 97% linear vs cyclic) through modulating the reaction conditions, thus demonstrating the advantage of an organocatalyzed method with this monomer family.


Polymerization of a coordinating monomer by O-ATRP with control over polymer backbone composition.

6. CONCLUSIONS AND FUTURE DIRECTIONS

In this review, we have attempted to summarize developments in O-ATRP for both new practitioners and veterans in the field. In the context of other metal-free CRPs, O-ATRP is an excellent method for the precise polymerization of acrylates, methacrylates, styrene, and vinylcyclopropanes using commercially available reagents. While a limited number of strongly reducing PCs are currently available from commercial sources, continued development in the design of such PCs is beginning to address this limitation. In addition, the application of reductive quenching PCs in O-ATRP, such as eosin Y and fluorescein, will help to address this issue given the greater availability of these compounds.

We have also attempted to provide a comprehensive review of the materials produced by this method, as well as its various applications in precision polymer synthesis. In particular, the tolerance of O-ATRP to a wide range of chemical functionalities has been one of its primary advantages, especially with respect to coordinating functionalities that are poorly tolerated in traditional metal-catalyzed ATRP methods. As a result, O-ATRP has found use in a wide range of applications, ranging from the synthesis of polymers with complex architectures through concurrent and orthogonal methods to drug delivery and biological imaging. Indeed, we are excited to see the many new applications in which this method will be used in the coming years.

Moving forward in the development of O-ATRP, we anticipate key challenges to overcome will include:

(1) Developing a better understanding of the mechanism of O-ATRP, both in the presence of oxidative and reductive quenching PCs. In both cases, developing a detailed understanding of the activation and deactivation processes will be crucial for identifying and overcoming limitations in the mechanism of O-ATRP. While the work highlighted in this review has begun to elucidate these mechanistic details, future work should focus on investigating a wider range of PCs and developing a more generalized understanding of the O-ATRP mechanism.

(2) Developing a general understanding of solvent effects in O-ATRP. Currently, knowledge surrounding solvent effects in this method is largely focused on dihydrophenazine PCs, because these are the catalysts that have mostly been employed in solvent effect studies thus far. However, similar solvent effects have not yet been reported for other PCs, possibly due to differences in their chemical structures and photophysical properties. As such, further work is necessary in this area to understand how solvent properties impact other PC families, with the ultimate goal being a generalized model to understand and predict solvent effects across a range of O-ATRP catalysts.

(3) Identifying important characteristics for successful initiators in O-ATRP. Aside from limited initiator screens and work investigating the impact of the halide identity (i.e., alkyl bromides vs chlorides), little is known about the initiator in O-ATRP. In the coming years, systematic investigations of O-ATRP initiators would be beneficial to understand how this reaction variable impacts control over the product polymer structure. In particular, if there are any special requirements for O-ATRP that are not present in other ATRP methods, identifying them would be of significant interest. In addition, because O-ATRP catalysts operate through outer sphere electron transfer (as opposed to inner sphere electron transfer as observed with many traditional ATRP catalysts), perhaps the use of new initiators may be possible in O-ATRP that cannot be employed in traditional ATRP methods.

(4) Expanding the monomer scope of O-ATRP, especially to monomers that are inaccessible by metal-catalyzed methods. While any advancements in the monomer scope of O-ATRP will certainly be useful to the field, it is our hope that future work will expand beyond applications that are already accessible through other methods to those that are currently inaccessible using ATRP. One example of such monomers could be α-olefins, which can be difficult to polymerize by ATRP due to the strong C─Br bond that forms at the end of the polymer chain during polymerization. Because of the strength of this bond, activation of the chain-end can be extremely challenging using traditional ATRP catalysts. However, it may be possible for some O-ATRP catalysts to reduce this bond, enabling the polymerization of this elusive class of monomers. As such, further investigation in this area is warranted.

(5) Developing new catalysts that are effective, easy to synthesize, and inexpensive. Currently, few catalysts exist that meet all three of these criteria. For example, many reductive quenching PCs (ex., xanthenes) are inexpensive and readily available for purchase, but the level of control they offer in O-ATRP is often lower than that available using oxidative quenching PCs (see , ). However, many oxidative quenching PCs (ex., phenoxazines) require multiple-step syntheses that can increase the barrier to using O-ATRP as a nonexpert. As such, the development of PCs that are highly effective, easy to synthesize or purchase, and inexpensive will have a significant impact on the field.

(6) Understanding the toxicity of O-ATRP catalysts, such that they can be employed in biological applications without the need for polymer purification. Since many of the catalysts employed in O-ATRP have only recently been developed, little information is known regarding their effects on biological organisms. Despite this fact, one of the most cited potential applications of O-ATRP is in biological applications. As such, research is critically needed to understand the biological effects of common O-ATRP PCs to ensure the compatibility of this method with sensitive biological systems.

(7) Expanding the range of irradiation wavelengths that can be employed in O-ATRP to longer wavelengths that can penetrate biological tissue, enabling in vivo applications of O-ATRP. Because longer wavelengths of light possess less energy than shorter wavelengths, thermodynamic restrictions may arise in which a single photon of light cannot impart enough energy to a PC for reduction of a C─Br bond. However, new photochemical processes employing two-photon excitations may be beneficial in this area, allowing a PC to harness the energy of multiple photons to access higher-energy excited states capable of activation in O-ATRP.

(8) The ability to synthesize high molecular weight polymers. Current O-ATRP methods typically produce polymers with Mn ~ 1–50 kDa, but polymers with Mn > 100 kDa are not uncommon using other polymerization techniques. Future work should seek to elucidate why this limit exists, as well as how it can be overcome.

(9) The ability to achieve high or quantitative monomer conversion. Especially with regards to the synthesis of block copolymers, achieving this property will be crucial for increasing the accessibility of this method.

By continuing to address these challenges, O-ATRP will continue to be established as a powerful metal-free strategy for the synthesis of precision polymers in a variety of advanced applications and fields.

ACKNOWLEDGMENTS

We are grateful for support from Colorado State University, the National Science Foundation Division of Chemistry (award 2055742), the Research Corporation for Science Advancement (Cottrell Scholar Award), and the National Institute of General Medical Sciences (award R35GM119702) of the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. D.A.C. thanks Cassidy Jackson for her support, as well as Max Kudisch, Cam Chrisman, Mariel Price, and Felipe Garcia Alzate for helpful discussion during the writing of this manuscript.

Biographies

•

Daniel A. Corbin was born in 1995 and raised in southeastern Virginia. He performed his undergraduate studies at James Madison University, where he obtained a B.S. in Chemistry with a minor in Mathematics in 2017. While there, he performed undergraduate research with Prof. Brycelyn Boardman, developing novel metallopolymers for photovoltaic applications. In 2017, Daniel moved to Colorado to pursue a Ph.D. in Chemistry under Prof. Garret Miyake at Colorado State University. His work in the Miyake group has primarily focused on studying key steps in the mechanism of O-ATRP, with the ultimate goal of improving catalyst design and expanding the utility of this method.

•

Garret M. Miyake grew up in Oregon and earned a B.S. at Pacific University. He performed Ph.D. studies with Eugene Chen at Colorado State University before conducting postdoctoral research with Robert Grubbs at the California Institute of Technology. He is currently an Associate Professor of Chemistry at Colorado State University. The Miyake group has research interests in the fields of photoredox catalysis, organocatalyzed atom-transfer radical polymerization, sustainable plastics, and the synthesis of block copolymers that self-assemble to photonic crystals. He has been recognized by the 2021 Journal of Polymer Science Innovation Award, the 2017 ACS Division of Polymer Chemistry Herman F. Mark Young Scholar Award, a Camille Dreyfus Teacher-Scholar Award, a Cottrell Scholar Award, and a Sloan Research Fellowship. He is also a cofounder of New Iridium and Cypris Materials.

Footnotes

The authors declare the following competing financial interest(s): G.M.M. is a cofounder of New Iridium, which is commercializing organic photoredox catalysts.

REFERENCES

(1) Matyjaszewski K; Xia JHAtom transfer radical polymerization. Chem. Rev2001, 101, 2921–2990. [PubMed] [Google Scholar]

(2) Braunecker WA; Matyjaszewski KControlled/living radical polymerization: features, developments, and perspectives. Prog. Polym. Sci2007, 32, 93–146. [Google Scholar]

(3) Matyjaszewski KAtom transfer radical polymerization (ATRP): current status and future perspectives. Macromolecules2012, 45, 4015–4039. [Google Scholar]

(4) Krys P; Matyjaszewski KKinetics of atom transfer radical polymerization. Eur. Polym. J2017, 89, 482–523. [Google Scholar]

(5) Matyjaszewski KAdvanced materials by atom transfer radical polymerization. Adv. Mater2018, 30, 1706441. [PubMed] [Google Scholar]

(6) Corrigan N; Jung K; Moad G; Hawker CJ; Matyjaszewski K; Boyer CReversible-deactivation radical polymerization (controlled/living radical polymerization): from discovery to materials design and applications. Prog. Polym. Sci2020, 111, 101311. [Google Scholar]

(7) Odian GGPrinciples of Polymerization. 4th ed.; Wiley: Hoboken, NJ, 2004. [Google Scholar]

(8) Wang JS; Matyjaszewski KControlled/”living” radical polymerization. Atom-transfer radical polymerization in the presence of transition-metal complexes. J. Am. Chem. Soc1995, 117, 5614–5615. [Google Scholar]

(9) Kato M; Kamigaito M; Sawamoto M; Higashimura TPolymerization of methyl-methacrylate with the carbon tetrachloride/dichlorotris(triphenylphosphine)ruthenium(II)/methylaluminum bis-(2,6-di-tert-butylphenoxide) initiating system: possibility of living radical polymerization. Macromolecules1995, 28, 1721–1723. [Google Scholar]

(10) Jakubowski W; Min K; Matyjaszewski KActivators regenerated by electron transfer for atom transfer radical polymerization of styrene. Macromolecules2006, 39, 39–45. [PubMed] [Google Scholar]

(11) Matyjaszewski K; Jakubowski W; Min K; Tang W; Huang JY; Braunecker WA; Tsarevsky NVDiminishing catalyst concentration in atom transfer radical polymerization with reducing agents. Proc. Natl. Acad. Sci. U. S. A2006, 103, 15309–15314. [PMC free article] [PubMed] [Google Scholar]

(12) Tasdelen MA; Uygun M; Yagci YPhotoinduced controlled radical polymerization in methanol. Macromol. Chem. Phys2010, 211, 2271–2275. [PubMed] [Google Scholar]

(13) Tasdelen MA; Uygun M; Yagci YPhotoinduced controlled radical polymerization. Macromol. Rapid Commun2011, 32, 58–62. [PubMed] [Google Scholar]

(14) Zhang T; Gieseler D; Jordan RLights on! A significant photoenhancement effect on ATRP by ambient laboratory light. Polym. Chem2016, 7, 775–779. [Google Scholar]

(15) Fors BP; Hawker CJControl of a living radical polymerization of methacrylates by light. Angew. Chem., Int. Ed2012, 51, 8850–8853. [PubMed] [Google Scholar]

(16) Magenau AJD; Strandwitz NC; Gennaro A; Matyjaszewski KElectrochemically mediated atom transfer radical polymerization. Science2011, 332, 81–84. [PubMed] [Google Scholar]

(17) Zhang G; Song IY; Ahn KH; Park T; Choi WFree radical polymerization initiated and controlled by visible light photocatalysis at ambient temperature. Macromolecules2011, 44, 7594–7599. [Google Scholar]

(18) Pan X; Tasdelen MA; Laun J; Junkers T; Yagci Y; Matyjaszewski KPhotomediated controlled radical polymerization. Prog. Polym. Sci2016, 62, 73–125. [Google Scholar]

(19) Chen M; Zhong MJ; Johnson JALight-controlled radical polymerization: mechanisms, methods, and applications. Chem. Rev2016, 116, 10167–10211. [PubMed] [Google Scholar]

(20) Dadashi-Silab S; Doran S; Yagci YPhotoinduced electron transfer reactions for macromolecular syntheses. Chem. Rev2016, 116, 10212–10275. [PubMed] [Google Scholar]

(21) Shanmugam S; Xu JT; Boyer CPhotocontrolled living polymerization systems with reversible deactivations through electron and energy transfer. Macromol. Rapid Commun2017, 38, 1700143. [PubMed] [Google Scholar]

(22) Pan XC; Fantin M; Yuan F; Matyjaszewski KExternally controlled atom transfer radical polymerization. Chem. Soc. Rev2018, 47, 5457–5490. [PubMed] [Google Scholar]

(23) Ludwig JR; Schindler CSCatalyst: sustainable catalysis. Chem. 2017, 2, 313–316. [Google Scholar]

(24) Ravelli D; Fagnoni M; Albini APhotoorganocatalysis. What for? Chem. Soc. Rev2013, 42, 97–113. [PubMed] [Google Scholar]

(25) Romero NA; Nicewicz DAOrganic photoredox catalysis. Chem. Rev2016, 116, 10075–10166. [PubMed] [Google Scholar]

(26) Shanmugam S; Boyer COrganic photocatalysts for cleaner polymer synthesis. Science2016, 352, 1053–1054. [PubMed] [Google Scholar]

(27) Trotta JT; Fors BPOrganic catalysts for photocontrolled polymerizations. Synlett2016, 27, 702–713. [Google Scholar]

(28) Pitre SP; McTiernan CD; Scaiano JCUnderstanding the kinetics and spectroscopy of photoredox catalysis and transition-metal-free alternatives. Acc. Chem. Res2016, 49, 1320–1330. [PubMed] [Google Scholar]

(29) Bonardi AH; Dumur F; Noirbent G; Lalevee J; Gigmes DOrganometallic vs organic photoredox catalysts for photocuring reactions in the visible region. Beilstein J. Org. Chem2018, 14, 3025–3046. [PMC free article] [PubMed] [Google Scholar]

(30) Bogdos MK; Pinard E; Murphy JAApplications of organocatalysed visible-light photoredox reactions for medicinal chemistry. Beilstein J. Org. Chem2018, 14, 2035–2064. [PMC free article] [PubMed] [Google Scholar]

(31) Corbin DA; Lim CH; Miyake GMPhenothiazines, dihydrophenazines, and phenoxazines: sustainable alternatives to precious-metal-based photoredox catalysts. Aldrichimica Acta2019, 52, 7–21. [PMC free article] [PubMed] [Google Scholar]

(32) Miyake GM; Theriot JCPerylene as an organic photocatalyst for the radical polymerization of functionalized vinyl monomers through oxidative quenching with alkyl bromides and visible light. Macromolecules2014, 47, 8255–8261. [Google Scholar]

(33) Treat NJ; Sprafke H; Kramer JW; Clark PG; Barton BE; Read de Alaniz J; Fors BP; Hawker CJMetal-free atom transfer radical polymerization. J. Am. Chem. Soc2014, 136, 16096–16101. [PubMed] [Google Scholar]

(34) Theriot JC; Lim CH; Yang H; Ryan MD; Musgrave CB; Miyake GMOrganocatalyzed atom transfer radical polymerization driven by visible light. Science2016, 352, 1082–1086. [PubMed] [Google Scholar]

(35) Pearson RM; Lim CH; McCarthy BG; Musgrave CB; Miyake GMOrganocatalyzed atom transfer radical polymerization using N-aryl phenoxazines as photoredox catalysts. J. Am. Chem. Soc2016, 138, 11399–11407. [PMC free article] [PubMed] [Google Scholar]

(36) Liu XD; Zhang LF; Cheng ZP; Zhu XLMetal-free photoinduced electron transfer-atom transfer radical polymerization (PET-ATRP) via a visible light organic photocatalyst. Polym. Chem2016, 7, 689–700. [Google Scholar]

(37) Theriot JC; McCarthy BG; Lim CH; Miyake GMOrganocatalyzed atom transfer radical polymerization: perspectives on catalyst design and performance. Macromol. Rapid Commun2017, 38, 1700040. [PMC free article] [PubMed] [Google Scholar]

(38) Discekici EH; Anastasaki A; Read de Alaniz J; Hawker CJEvolution and future directions of metal-free atom transfer radical polymerization. Macromolecules2018, 51, 7421–7434. [Google Scholar]

(39) Yilmaz G; Yagci YPhotoinduced metal-free atom transfer radical polymerizations: state-of-the-art, mechanistic aspects and applications. Polym. Chem2018, 9, 1757–1762. [Google Scholar]

(40) Yilmaz G; Kutahya C; Allushi A; Aydogan C; Aykac S; Yagci Y Photoinduced metal free strategies for atom transfer radical polymerization. In Reversible Deactivation Radical Polymerization: Mechanisms and Synthetic Methodologies; ACS Symposium Series; American Chemical Society, 2018; Vol. 1284, pp 263–271.. [Google Scholar]

(41) Kreutzer J; Yagci YMetal free reversible-deactivation radical polymerizations: advances, challenges, and opportunities. Polymers2018, 10, 35. [PMC free article] [PubMed] [Google Scholar]

(42) Zhu N; Hu X; Fang Z; Guo KContinuous flow photoinduced reversible deactivation radical polymerization. Chemphotochem2018, 2, 831–838. [Google Scholar]

(43) Gong HH; Gu Y; Chen MControlled/living radical polymerization of semifluorinated (meth)acrylates. Synlett2018, 29, 1543–1551. [Google Scholar]

(44) Hu X; Zhu N; Guo KAdvances in organocatalyzed atom transfer radical polymerization. Adv. Polym. Technol2019, 2019, 7971683. [Google Scholar]

(45) Li N; Hu X; Fang L; Kou JH; Ni YR; Lu CHOrganocatalyzed atom transfer radical polymerization. Prog. Chem2019, 31, 791–799. [Google Scholar]

(46) Aklujkar PS; Rao ARDevelopments in the components of metal-free photoinitiated organocatalyzed-atom transfer radical polymerization (O-ATRP). Chemistryselect2020, 5, 14884–14899. [Google Scholar]

(47) Chiefari J; Chong YK; Ercole F; Krstina J; Jeffery J; Le TPT; Mayadunne RTA; Meijs GF; Moad CL; Moad G; et al. Living free-radical polymerization by reversible addition-fragmentation chain transfer: The RAFT process. Macromolecules1998, 31, 5559–5562. [Google Scholar]

(48) Xu JT; Jung K; Atme A; Shanmugam S; Boyer CA robust and versatile photoinduced living polymerization of conjugated and unconjugated monomers and its oxygen tolerance. J. Am. Chem. Soc2014, 136, 5508–5519. [PubMed] [Google Scholar]

(49) Xu JT; Jung K; Boyer COxygen tolerance study of photoinduced electron transfer-reversible addition-fragmentation chain transfer (PET-RAFT) polymerization mediated by Ru(bpy)3Cl2. Macromolecules2014, 47, 4217–4229. [Google Scholar]

(50) Xu JT; Shanmugam S; Boyer COrganic electron donor-acceptor photoredox catalysts: enhanced catalytic efficiency toward controlled radical polymerization. ACS Macro Lett. 2015, 4, 926–932. [PubMed] [Google Scholar]

(51) Xu JT; Shanmugam S; Duong HT; Boyer COrgano-photocatalysts for photoinduced electron transfer-reversible addition-fragmentation chain transfer (PET-RAFT) polymerization. Polym. Chem2015, 6, 5615–5624. [Google Scholar]

(52) Phommalysack-Lovan J; Chu YY; Boyer C; Xu JTPET-RAFT polymerisation: towards green and precision polymer manufacturing. Chem. Commun2018, 54, 6591–6606. [PubMed] [Google Scholar]

(53) Otsu T; Yoshida MRole of initiator-transfer agent-terminator (iniferter) in radical polymerizations: polymer design by organic disulfides as iniferters. Makromol. Chem., Rapid Commun1982, 3, 127–132. [Google Scholar]

(54) Otsu TIniferter concept and living radical polymerization. J. Polym. Sci., Part A: Polym. Chem2000, 38, 2121–2136. [Google Scholar]

(55) Quinn JF; Barner L; Barner-Kowollik C; Rizzardo E; Davis TPReversible addition-fragmentation chain transfer polymerization initiated with ultraviolet radiation. Macromolecules2002, 35, 7620–7627. [Google Scholar]

(56) Grubbs RBNitroxide-mediated radical polymerization: limitations and versatility. Polym. Rev2011, 51, 104–137. [Google Scholar]

(57) Nicolas J; Guillaneuf Y; Lefay C; Bertin D; Gigmes D; Charleux BNitroxide-mediated polymerization. Prog. Polym. Sci2013, 38, 63–235. [Google Scholar]

(58) Audran G; Bagryanskaya EG; Marque SRA; Postnikov PNew variants of nitroxide mediated polymerization. Polymers2020, 12, 1481. [PMC free article] [PubMed] [Google Scholar]

(59) Yoshida EPhoto-living radical polymerization of methyl methacrylate by a nitroxide mediator. Colloid Polym. Sci2008, 286, 1663–1666. [Google Scholar]

(60) Guillaneuf Y; Bertin D; Gigmes D; Versace DL; Lalevee J; Fouassier JPToward nitroxide-medlated photopolymerization. Macromolecules2010, 43, 2204–2212. [Google Scholar]

(61) Ni YY; Zhang LF; Cheng ZP; Zhu XLIodine-mediated reversible-deactivation radical polymerization: a powerful strategy for polymer synthesis. Polym. Chem2019, 10, 2504–2515. [Google Scholar]

(62) Tatemoto M; Yutani Y; Fujiwara KNovel Iodine-Containing Compound and Preparation Thereof. European Patent EP 0272698 A2, 1988.

(63) Matyjaszewski K; Gaynor S; Wang JSControlled radical polymerizations: the use of alkyl iodides in degenerative transfer. Macromolecules1995, 28, 2093–2095. [Google Scholar]

(64) Isse AA; Lin CY; Coote ML; Gennaro AEstimation of standard reduction potentials of halogen atoms and alkyl halides. J. Phys. Chem. B2011, 115, 678–684. [PubMed] [Google Scholar]

(65) Buss BL; Lim C-H; Miyake GMDimethyl dihydroacridines as photocatalysts in organocatalyzed atom transfer radical polymerization of acrylate monomers. Angew. Chem., Int. Ed2020, 59, 3209–3217. [PMC free article] [PubMed] [Google Scholar]

(66) Liu YM; Chen Q; Tong YJ; Ma YG9,9-Dimethyl dihydroacridine-based organic photocatalyst for atom transfer radical polymerization from modifying “unstable” electron donor. Macromolecules2020, 53, 7053–7062. [Google Scholar]

(67) Shaw MH; Twilton J; MacMillan DWCPhotoredox catalysis in organic chemistry. J. Org. Chem2016, 81, 6898–6926. [PMC free article] [PubMed] [Google Scholar]

(68) Prier CK; Rankic DA; MacMillan DWCVisible light photoredox catalysis with transition metal complexes: applications in organic synthesis. Chem. Rev2013, 113, 5322–5363. [PMC free article] [PubMed] [Google Scholar]

(73) Furst L; Matsuura BS; Narayanam JMR; Tucker JW; Stephenson CRJVisible light-mediated intermolecular c-h functionalization of electron-rich heterocycles with malonates. Org. Lett2010, 12, 3104–3107. [PubMed] [Google Scholar]

(74) Bachman JC; Kavian R; Graham DJ; Kim DY; Noda S; Nocera DG; Shao-Horn Y; Lee SWElectrochemical polymerization of pyrene derivatives on functionalized carbon nanotubes for pseudocapacitive electrodes. Nat. Commun2015, 6, 7040. [PMC free article] [PubMed] [Google Scholar]

(75) Singh-Rachford TN; Castellano FNTriplet sensitized red-to-blue photon upconversion. J. Phys. Chem. Lett2010, 1, 195–200. [Google Scholar]

(76) Grimme S; Parac MSubstantial errors from time-dependent density functional theory for the calculation of excited states of large pi systems. ChemPhysChem2003, 4 (3), 292–295. [PubMed] [Google Scholar]

(77) Kikuchi K; Niwa T; Takahashi Y; Ikeda H; Miyashi TQuenching mechanism in a highly exothermic region of the rehmweller relationship for electron-transfer fluorescence quenching. J. Phys. Chem1993, 97, 5070–5073. [Google Scholar]

(78) Theriot JC; McCarthy BG; Lim CH; Miyake GMOrganocatalyzed atom transfer radical polymerization: perspectives on catalyst design and performance. Macromol. Rapid Commun2017, 38, 1700040. [PMC free article] [PubMed] [Google Scholar]

(79) Pan XC; Fang C; Fantin M; Malhotra N; So WY; Peteanu LA; Isse AA; Gennaro A; Liu P; Matyjaszewski KMechanism of photoinduced metal-free atom transfer radical polymerization: experimental and computational studies. J. Am. Chem. Soc2016, 138, 2411–2425. [PubMed] [Google Scholar]

(80) Zhao YC; Gong HH; Jiang KM; Yan SJ; Lin J; Chen MOrganocatalyzed photoredox polymerization from aromatic sulfonyl halides: facilitating graft from aromatic C-H bonds. Macromolecules2018, 51, 938–946. [Google Scholar]

(81) Cole JP; Federico CR; Lim CH; Miyake GMPhotoinduced organocatalyzed atom transfer radical polymerization using low ppm catalyst loading. Macromolecules2019, 52, 747–754. [PMC free article] [PubMed] [Google Scholar]

(82) Luo J; Zhang JDonor–acceptor fluorophores for visible-light-promoted organic synthesis: photoredox/Ni dual catalytic C(sp3)–C(sp2) cross-coupling. ACS Catal. 2016, 6, 873–877. [Google Scholar]

(83) Shen T; Zhao ZG; Yu Q; Xu HJPhotosensitized reduction of benzil by heteroatom-containing anthracene dyes. J. Photochem. Photobiol., A1989, 47, 203–212. [Google Scholar]

(84) Timpe HJ; Kronfeld KPLight-induced polymer and polymerization reactions. 33. direct photoinitiation of methylmethacrylate polymerization by excited-states of ketones. J. Photochem. Photobiol., A1989, 46, 253–267. [Google Scholar]

(85) Timpe HJ; Kronfeld KP; Lammel U; Fouassier JP; Lougnot DJExcited-states of ketones as electron-donors ketone iodonium salt systems as photoinitiators for radical polymerization. J. Photochem. Photobiol., A1990, 52, 111–122. [Google Scholar]

(86) Arias-Rotondo DM; McCusker JKThe photophysics of photoredox catalysis: a roadmap for catalyst design. Chem. Soc. Rev2016, 45, 5803–5820. [PubMed] [Google Scholar]

(87) Su X; Jessop PG; Cunningham MFVersatility of organocatalyzed atom transfer radical polymerization and CO2− switching for preparing both hydrophobic and hydrophilic polymers with the recycling of a photocatalyst. Macromolecules2019, 52, 6725–6733. [Google Scholar]

(88) McCarthy B; Sartor S; Cole J; Damrauer N; Miyake GMSolvent effects and side reactions in organocatalyzed atom transfer radical polymerization for enabling the controlled polymerization of acrylates catalyzed by diaryl dihydrophenazines. Macromolecules2020, 53, 9208–9219. [PMC free article] [PubMed] [Google Scholar]

(89) Corbin DA; Puffer KO; Chism KA; Cole JP; Theriot JC; McCarthy BG; Buss BL; Lim C-H; Lincoln SR; Newell BS; et al. Radical addition to N,N-diaryl dihydrophenazine photoredox catalysts and implications in photoinduced organocatalyzed atom transfer radical polymerization. Macromolecules2021, 54, 4507–4516. [PMC free article] [PubMed] [Google Scholar]

(90) Zhang Y; Jiang D; Fang Z; Zhu N; Sun N; He W; Liu C; Zhao L; Guo KPhotomediated core modification of organic photoredox catalysts in radical addition: mechanism and applications. Chem. Sci2021, 12, 9432. [PMC free article] [PubMed] [Google Scholar]

(91) Pan XC; Lamson M; Yan JJ; Matyjaszewski KPhotoinduced metal-free atom transfer radical polymerization of acrylonitrile. ACS Macro Lett. 2015, 4, 192–196. [PubMed] [Google Scholar]

(92) Huang WJ; Zhai JL; Hu X; Duan JD; Fang Z; Zhu N; Guo KContinuous flow photoinduced phenothiazine derivatives catalyzed atom transfer radical polymerization. Eur. Polym. J2020, 126, 109565. [Google Scholar]

(93) Tran HM; Phan LNT; Le TV; Truong TT; Nguyen TH; Truong KT; Nguyen LTT; Phong MT; Nguyen HTPhenothiazine derivative as organic photocatalyst for metal free atom transfer radical polymerization. Polymer2019, 43, 496–502. [Google Scholar]

(94) Nguyen TH; Nguyen LTT; Nguyen VQ; Phan LNT; Zhang G; Yokozawa T; Phung DTT; Nguyen HTSynthesis of poly(3-hexylthiophene) based rod-coil conjugated block copolymers via photoinduced metal-free atom transfer radical polymerization. Polym. Chem2018, 9, 2484–2493. [Google Scholar]

(95) McCarthy BG; Pearson RM; Lim CH; Sartor SM; Damrauer NH; Miyake GMStructure-property relationships for tailoring phenoxazines as reducing photoredox catalysts. J. Am. Chem. Soc2018, 140, 5088–5101. [PMC free article] [PubMed] [Google Scholar]

(96) Park GS; Back J; Choi EM; Lee E; Son KSVisible light-mediated metal-free atom transfer radical polymerization with N-trifluoromethylphenyl phenoxazines. Eur. Polym. J2019, 117, 347–352. [Google Scholar]

(97) Bian C; Zhou YN; Deetz JD; Luo ZHExperimental and computational investigation of oxidative quenching governed aqueous organocatalyzed atom transfer radical polymerization. Chem. Eng. J2019, 362, 721–730. [Google Scholar]

(98) Zhang Z; Chen W; Zhang Y; Wang Y; Tian Y; Fang L; Ba XPhotoredox organocatalysts with thermally activated delayed fluorescence for visible-light-driven atom transfer radical polymerization. Macromolecules2021, 54, 4633–4640. [Google Scholar]

(99) McCarthy B; Miyake GMOrganocatalyzed Atom transfer radical polymerization catalyzed by core modified N-aryl phenoxazines performed under air. ACS Macro Lett. 2018, 7, 1016–1021. [PMC free article] [PubMed] [Google Scholar]

(100) Flory PJPrinciples of Polymer Chemistry; Cornell University Press: New York, 1995. [Google Scholar]

(101) Turro NJ; Ramamurthy V; Scaiano JCModern Molecular Photochemistry of Organic Molecules; University Science Books: Mill Valley, CA, 2010. [Google Scholar]

(102) Beuermann S; Buback MRate coefficients of free-radical polymerization deduced from pulsed laser experiments. Prog. Polym. Sci2002, 27, 191–254. [Google Scholar]

(103) Gillies MB; Matyjaszewski K; Norrby PO; Pintauer T; Poli R; Richard PA DFT study of R-X bond dissociation enthalpies of relevance to the initiation process of atom transfer radical polymerization. Macromolecules2003, 36, 8551–8559. [Google Scholar]

(104) Tang W; Matyjaszewski KEffects of initiator structure on activation rate constants in ATRP. Macromolecules2007, 40, 1858–1863. [Google Scholar]

(105) Hartman RL; McMullen JP; Jensen KFDeciding whether to go with the flow: evaluating the merits of flow reactors for synthesis. Angew. Chem., Int. Ed2011, 50, 7502–7519. [PubMed] [Google Scholar]

(106) Tucker JW; Zhang Y; Jamison TF; Stephenson CRJVisible-light photoredox catalysis in flow. Angew. Chem., Int. Ed2012, 51, 4144–4147. [PMC free article] [PubMed] [Google Scholar]

(107) Junkers T; Wenn BContinuous photoflow synthesis of precision polymers. React. Chem. Eng2016, 1, 60–64. [Google Scholar]

(108) Buss BL; Miyake GMPhotoinduced controlled radical polymerizations performed in flow: methods, products, and opportunities. Chem. Mater2018, 30, 3931–3942. [PMC free article] [PubMed] [Google Scholar]

(109) Allushi A; Jockusch S; Yilmaz G; Yagci YPhotoinitiated metal-free controlled/living radical polymerization using polynuclear aromatic hydrocarbons. Macromolecules2016, 49, 7785–7792. [Google Scholar]

(110) Wang GX; Lu M; Zhou MJ; Liang EX; He BHPhotoinduced ATRP of MMA under blue light irradiation in the presence of 3,4,9,10-tetra-(12-alkoxycarbonyl)-perylene as a photocatalyst. Iran. Polym. J2018, 27, 43–48. [Google Scholar]

(111) Zeng LL; Xie WY; Yang CX; Liang EX; Wang GXPhotomediated atom transfer radical polymerization of MMA under long-wavelength light irradiation. Iran. Polym. J2018, 27, 881–887. [Google Scholar]

(112) Ma Q; Song JS; Zhang X; Jiang Y; Ji L; Liao SHMetal-free atom transfer radical polymerization with ppm catalyst loading under sunlight. Nat. Commun2021, 12, 429. [PMC free article] [PubMed] [Google Scholar]

(113) Kutahya C; Allushi A; Isci R; Kreutzer J; Ozturk T; Yilmaz G; Yagci YPhotoinduced metal-free atom transfer radical polymerization using highly conjugated thienothiophene derivatives. Macromolecules2017, 50, 6903–6910. [Google Scholar]

(114) Huang ZC; Gu Y; Liu XD; Zhang LF; Cheng ZP; Zhu XLMetal-free atom transfer radical polymerization of methyl methacrylate with ppm level of organic photocatalyst. Macromol. Rapid Commun2017, 38, 1600461. [PubMed] [Google Scholar]

(115) Singh VK; Yu C; Badgujar S; Kim Y; Kwon Y; Kim D; Lee J; Akhter T; Thangavel G; Park LS; et al. Highly efficient organic photocatalysts discovered via a computer-aided-design strategy for visible-light-driven atom transfer radical polymerization. Nat. Catal2018, 1, 794–804. [Google Scholar]

(116) Wang GX; Lu M; Zhou MJPhoto-induced metal-free ATRP of MMA with 2,7-bi-(N-penothiazinyl)fluorenone as photocatalyst. J. Macromol. Sci., Part A: Pure Appl.Chem2018, 55, 85–89. [Google Scholar]

(117) Wang L; Li R; Zhang KAIAtom transfer radical polymerization (ATRP) catalyzed by visible light-absorbed small molecule organic semiconductors. Macromol. Rapid Commun2018, 39, 1800466. [PubMed] [Google Scholar]

(118) Ma Q; Zhang X; Ji L; Liao SHBINOLs as visible light photocatalysts for metal-free atom transfer radical polymerization. Polym. Chem2019, 10, 6662–6668. [Google Scholar]

(119) Jia T; Huang S; Bohra H; Wang MFExamining derivatives of quinacridone, diketopyrrolopyrrole and indigo as the visible-light organic photocatalysts for metal-free atom transfer radical polymerization. Dyes Pigm. 2019, 165, 223–230. [Google Scholar]

(120) Yang L; Huang YJ; Peng YT; Liu F; Zhang C; He HC; Wang J; Jiang L; Zhou YPyridine-diketopyrrolopyrrole-based novel metal-free visible-light organophotoredox catalyst for atom-transfer radical polymerization. J. Phys. Chem. A2020, 124, 1068–1075. [PubMed] [Google Scholar]

(121) Li MM; Wang SX; Li FF; Zhou L; Lei LOrganocatalyzed atom transfer radical polymerization (ATRP) using triarylsulfonium hexafluorophosphate salt (THS) as a photocatalyst. Polym. Chem2020, 11, 2222–2229. [Google Scholar]

(122) Ma W; Zhang X; Ma Y; Chen D; Wang L; Zhao C; Yang WPhotoinduced controlled radical polymerization of methacrylates with benzaldehyde derivatives as organic catalysts. Polym. Chem2017, 8, 3574–3585. [Google Scholar]

(123) Zaborniak I; Chmielarz P; Wolski KRiboflavin-induced metal-free ATRP of (meth)acrylates. Eur. Polym. J2020, 140, 110055. [Google Scholar]

(124) Kutahya C; Aykac FS; Yilmaz G; Yagci YLED and visible light-induced metal free ATRP using reducible dyes in the presence of amines. Polym. Chem2016, 7, 6094–6098. [Google Scholar]

(125) Wang YA; Li HL; Dong JH; Hu LJ; Wei DL; Bai LJ; Yang HW; Chen HRecyclable bio-based photoredox catalyst in metal-free atom transfer radical polymerization. Macromol. Chem. Phys2021, 222, 2000406. [Google Scholar]

(126) Xu XM; Xu X; Zeng YN; Zhang FAOxygen-tolerant photo-induced metal-free atom transfer radical polymerization. J. Photochem. Photobiol., A2021, 411, 113191. [Google Scholar]

(127) Allushi A; Kutahya C; Aydogan C; Kreutzer J; Yilmaz G; Yagci YConventional type II photoinitiators as activators for photoinduced metal-free atom transfer radical polymerization. Polym. Chem2017, 8, 1972–1977. [Google Scholar]

(128) Liu LC; Yi YSPhoto-mediated metal free atom transfer radical polymerization of acrylamide in water. J. Appl. Polym. Sci2018, 135, 46567. [Google Scholar]

(129) Zaborniak I; Macior A; Chmielarz P; Caceres Najarro M; Iruthayaraj JLignin-based thermoresponsive macromolecules via vitamin-induced metal-free ATRP. Polymer2021, 219, 123537. [Google Scholar]

(130) Anslyn EV; Dougherty DAModern Physical Organic Chemistry. University Science: Mill Valley, CA, 2006. [Google Scholar]

(131) Lakowicz JRPrinciples of Fluorescence Spectroscopy, 3rd ed.; Springer: New York, 2006. [Google Scholar]

(132) Turro NJModern Molecular Photochemistry; University Science Books, 1991. [Google Scholar]

(133) Kasha MCharacterization of electronic transitions in complex molecules. Discuss. Faraday Soc1950, 9, 14–19. [Google Scholar]

(134) Arnaut LG; Formosinho SJ; Burrows HChemical Kinetics: From Molecular Structure to Chemical Reactivity, 1st ed.; Elsevier: Amsterdam, 2007. [Google Scholar]

(135) Lim CH; Ryan MD; McCarthy BG; Theriot JC; Sartor SM; Damrauer NH; Musgrave CB; Miyake GMIntramolecular charge transfer and ion pairing in N,N-diaryl dihydrophenazine photoredox catalysts for efficient organocatalyzed atom transfer radical polymerization. J. Am. Chem. Soc2017, 139, 348–355. [PMC free article] [PubMed] [Google Scholar]

(136) Sartor SM; McCarthy BG; Pearson RM; Miyake GM; Damrauer NHExploiting charge-transfer states for maximizing intersystem crossing yields in organic photoredox catalysts. J. Am. Chem. Soc2018, 140, 4778–4781. [PMC free article] [PubMed] [Google Scholar]

(137) Sartor SM; Lattke YM; McCarthy BG; Miyake GM; Damrauer NHEffects of naphthyl connectivity on the photophysics of compact organic charge-transfer photoredox catalysts. J. Phys. Chem. A2019, 123, 4727–4736. [PMC free article] [PubMed] [Google Scholar]

(138) Sartor SM; Chrisman CH; Pearson RM; Miyake GM; Damrauer NHDesigning high-triplet-yield phenothiazine donor-acceptor complexes for photoredox catalysis. J. Phys. Chem. A2020, 124, 817–823. [PubMed] [Google Scholar]

(139) Ryan MD; Pearson RM; French TA; Miyake GMImpact of light intensity on control in photoinduced organocatalyzed atom transfer radical polymerization. Macromolecules2017, 50, 4616–4622. [PMC free article] [PubMed] [Google Scholar]

(140) Dolinski ND; Page ZA; Discekici EH; Meis D; Lee IH; Jones GR; Whitfield R; Pan XC; McCarthy BG; Shanmugam S; et al. What happens in the dark? Assessing the temporal control of photo-mediated controlled radical polymerizations. J. Polym. Sci., Part A: Polym. Chem2019, 57, 268–273. [PMC free article] [PubMed] [Google Scholar]

(141) Guo JK; Luo ZHHow the catalyst circulates and works in organocatalyzed atom transfer radical polymerization. AIChE J. 2018, 64, 2581–2591. [Google Scholar]

(142) Jockusch S; Yagci YThe active role of excited states of phenothiazines in photoinduced metal free atom transfer radical polymerization: singlet or triplet excited states? Polym. Chem2016, 7, 6039–6043. [Google Scholar]

(143) Koyama D; Dale HJA; Orr-Ewing AJUltrafast observation of a photoredox reaction mechanism: photoinitiation in organocatalyzed atom-transfer radical polymerization. J. Am. Chem. Soc2018, 140, 1285–1293. [PubMed] [Google Scholar]

(144) Lewis-Borrell L; Sneha M; Bhattacherjee A; Clark IP; Orr-Ewing AJMapping the multi-step mechanism of a photoredox catalyzed atom-transfer radical polymerization reaction by direct observation of the reactive intermediates. Chem. Sci2020, 11, 4475–4481. [PMC free article] [PubMed] [Google Scholar]

(145) Bhattacherjee A; Sneha M; Lewis-Borrell L; Amoruso G; Oliver TAA; Tyler J; Clark IP; Orr-Ewing AJSinglet and triplet contributions to the excited-state activities of dihydrophenazine, phenoxazine, and phenothiazine organocatalysts used in atom transfer radical polymerization. J. Am. Chem. Soc2021, 143, 3613–3627. [PubMed] [Google Scholar]

(146) Lattke YM; Corbin DA; Sartor SM; McCarthy BG; Miyake GM; Damrauer NHInterrogation of O-ATRP activation conducted by singlet and triplet excited states of phenoxazine photocatalysts. J. Phys. Chem. A2021, 125, 3109–3121. [PMC free article] [PubMed] [Google Scholar]

(147) Du Y; Pearson RM; Lim CH; Sartor SM; Ryan MD; Yang HS; Damrauer NH; Miyake GMStrongly reducing, visible-light organic photoredox catalysts as sustainable alternatives to precious metals. Chem. - Eur. J2017, 23, 10962–10968. [PMC free article] [PubMed] [Google Scholar]

(148) Ryan MD; Theriot JC; Lim CH; Yang HS; Lockwood AG; Garrison NG; Lincoln SR; Musgrave CB; Miyake GMSolvent effects on the intramolecular charge transfer character of N,N-diaryl dihydrophenazine catalysts for organocatalyzed atom transfer radical polymerization. J. Polym. Sci., Part A: Polym. Chem2017, 55, 3017–3027. [PMC free article] [PubMed] [Google Scholar]

(149) Krzysztof M; Thomas PDHandbook of Radical Polymerization; Wiley-Interscience: Hoboken, N.J., 2002. [Google Scholar]

(150) Moad G; Solomon DH 5-Termination. In The Chemistry of Radical Polymerization, 2nd ed.; Moad G, Solomon DH, Eds.; Elsevier Science Ltd: Amsterdam, 2005; pp 233–278. [Google Scholar]

(151) Corbin DA; McCarthy BG; van de Lindt Z; Miyake GMRadical cations of phenoxazine and dihydrophenazine photoredox catalysts and their role as deactivators in organocatalyzed atom transfer radical polymerization. Macromolecules2021, 54, 4726–4738. [PMC free article] [PubMed] [Google Scholar]

(152) Kotsinaris A; Kyriacou G; Lambrou CElectrochemical reduction of dichloromethane to higher hydrocarbons. J. Appl. Electrochem1998, 28, 613–616. [Google Scholar]

(153) Bian C; Zhou YN; Guo JK; Luo ZHAqueous metal-free atom transfer radical polymerization: experiments and model-based approach for mechanistic understanding. Macromolecules2018, 51, 2367–2376. [Google Scholar]

(154) Dadashi-Silab S; Pan XC; Matyjaszewski KPhenyl benzo-[b]-phenothiazine as a visible light photoredox catalyst for metal-free atom transfer radical polymerization. Chem. - Eur. J2017, 23, 5972–5977. [PubMed] [Google Scholar]

(155) Ramsey BL; Pearson RM; Beck LR; Miyake GMPhotoinduced organocatalyzed atom transfer radical polymerization using continuous flow. Macromolecules2017, 50, 2668–2674. [PMC free article] [PubMed] [Google Scholar]

(156) Beuermann S; Buback M; Hesse P; Lacik IFree-radical propagation rate coefficient of nonionized methacrylic acid in aqueous solution from low monomer concentrations to bulk polymerization. Macromolecules2006, 39, 184–193. [Google Scholar]

(157) Beuermann S; Buback M; Davis TP; Gilbert RG; Hutchinson RA; Olaj OF; Russell GT; Schweer J; vanHerk AMCritically evaluated rate coefficients for free-radical polymerization. 2. propagation rate coefficients for methyl methacrylate. Macromol. Chem. Phys1997, 198, 1545–1560. [Google Scholar]

(158) Beuermann S; Buback M; Davis TP; Gilbert RG; Hutchinson RA; Kajiwara A; Klumperman B; Russell GTCritically evaluated rate coefficients for free-radical polymerization. 3.propagation rate coefficients for alkyl methacrylates. Macromol. Chem. Phys2000, 201, 1355–1364. [Google Scholar]

(159) Hutchinson RA; Beuermann S; Paquet DA; McMinn JHDetermination of free-radical propagation rate coefficients for alkyl methacrylates by pulsed-laser polymerization. Macromolecules1997, 30, 3490–3493. [Google Scholar]

(160) Buback M; Kurz CHFree-radical propagation rate coefficients for cyclohexyl methacrylate, glycidyl methacrylate and 2-hydroxyethyl methacrylate homopolymerizations. Macromol. Chem. Phys1998, 199, 2301–2310. [Google Scholar]

(161) Hutchinson RA; Beuermann S; Paquet DA; McMinn JH; Jackson CDetermination of free-radical propagation rate coefficients for cycloalkyl and functional methacrylates by pulsed-laser polymerization. Macromolecules1998, 31, 1542–1547. [Google Scholar]

(162) Zammit MD; Coote ML; Davis TP; Willett GDEffect of the ester side-chain on the propagation kinetics of alkyl methacrylates:an entropic or enthalpic effect? Macromolecules1998, 31, 955–963. [Google Scholar]

(163) Manders LGPulsed Initiation Polymerization: Applications in Homogeneous and Heterogeneous Radical Systems; Technische Universiteit Eindhoven: Eindhoven, 1997. [Google Scholar]

(164) Lyons RA; Hutovic J; Piton MC; Christie DI; Clay PA; Manders BG; Kable SH; Gilbert RGPulsed-laser polymerization measurements of the propagation rate coefficient for butyl acrylate. Macromolecules1996, 29, 1918–1927. [Google Scholar]

(165) Buback M; Kurz CH; Schmaltz CPressure dependence of propagation rate coefficients in free-radical homopolymerizations of methyl acrylate and dodecyl acrylate. Macromol. Chem. Phys1998, 199, 1721–1727. [Google Scholar]

(166) Buback M; Gilbert RG; Hutchinson RA; Klumperman B; Kuchta FD; Manders BG; O'Driscoll KF; Russell GT; Schweer JCritically evaluated rate coefficients for free-radical polymerization. 1. propagation rate coefficient for styrene. Macromol. Chem. Phys1995, 196, 3267–3280. [Google Scholar]

(167) International Workshop on Polymer Reaction, E. Reichert KH; Moritz HU; Dechema In 5th International Workshop on Polymer Reaction Engineering: Papers of the 5th International Workshop on Polymer Reaction Engineering, Berlin, 9–11 October, 1995; DECHEMA Monographs 131; VCH Publishers: Frankfurt am Main, New York, 1995. [Google Scholar]

(168) Deibert S; Bandermann F; Schweer J; Sarnecki JPropagation rate coefficient of free-radical polymerization of 1,3-butadiene. Makromol. Chem., Rapid Commun1992, 13, 351–355. [Google Scholar]

(169) Hutchinson RA; Aronson MT; Richards JRAnalysis of pulsed-laser-generated molecular-weight distributions for the determination of propagation rate coefficients. Macromolecules1993, 26, 6410–6415. [Google Scholar]

(170) Ramakers G; Krivcov A; Trouillet V; Welle A; Mobius H; Junkers TOrganocatalyzed photo-atom transfer radical polymerization of methacrylic acid in continuous flow and surface grafting. Macromol. Rapid Commun2017, 38, 1700423. [PubMed] [Google Scholar]

(171) Kim YA; Park GS; Son KSAmphiphilic diblock and crosslinked copolymers synthesized via metal-free atom transfer radical polymerization. Polym. Int2018, 67, 127–131. [Google Scholar]

(172) Wang JF; Yuan L; Wang ZK; Rahman MA; Huang YC; Zhu TY; Wang RB; Cheng JJ; Wang CP; Chu FX; et al. Photoinduced metal-free atom transfer radical polymerization of biomass-based monomers. Macromolecules2016, 49, 7709–7717. [Google Scholar]

(173) Ni BQ; Wang DK; Zhang H; Li PQ; Niu TFVisible light controlled polymerization of azide-derived monomers: a facile, metal-free PET-ATRP route to construct azide polymers. Macromol. Chem. Phys2019, 220, 1800529. [Google Scholar]

(174) Mehmood R; Mofarah SS; Rawal A; Tomasetig F; Wang XC; Yang JL; Koshy P; Sorrell CCGreen synthesis of zwitterion-functionalized nano-octahedral ceria for enhanced intracellular delivery and cancer therapy. ACS Sustainable Chem. Eng2019, 7, 9189–9201. [Google Scholar]

(175) Aydogan C; Yilmaz G; Yagci YSynthesis of hyperbranched polymers by photoinduced metal-free ATRP. Macromolecules2017, 50, 9115–9120. [Google Scholar]

(176) Ramakers G; Wackers G; Trouillet V; Welle A; Wagner P; Junkers TLaser-grafted molecularly imprinted polymers for the detection of histamine from organocatalyzed atom transfer radical polymerization. Macromolecules2019, 52, 2304–2313. [Google Scholar]

(177) Narupai B; Page ZA; Treat NJ; McGrath AJ; Pester CW; Discekici EH; Dolinski ND; Meyers GF; Read de Alaniz J; Hawker CJSimultaneous preparation of multiple polymer brushes under ambient conditions using microliter volumes. Angew. Chem., Int. Ed2018, 57, 13433–13438. [PubMed] [Google Scholar]

(178) Liu QR; Jian LH; Liu RQ; Yang HX; Kong JM; Zhang XJMetal-free photoinduced atom transfer radical polymerization for highly sensitive detection of lung cancer DNA. Chem. - Eur. J2020, 26, 1633–1639. [PubMed] [Google Scholar]

(179) Li SP; Mohamed AI; Pande V; Wang H; Cuthbert J; Pan XC; He HK; Wang ZY; Viswanathan V; Whitacre JF; et al. Single-ion homopolymer electrolytes with high transference number prepared by click chemistry and photoinduced metal-free atom-transfer radical polymerization. ACS Energy Lett. 2018, 3, 20–27. [Google Scholar]

(180) Zeng GJ; Liu MY; Shi KX; Heng CN; Mao LC; Wan Q; Huang HY; Deng FJ; Zhang XY; Wei YSurface modification of nanodiamond through metal free atom transfer radical polymerization. Appl. Surf. Sci2016, 390, 710–717. [Google Scholar]

(181) Luo YRHandbook of Bond Dissociation Energies in Organic Compounds; CRC Press: Boca Raton, FL, 2002. [Google Scholar]

(182) Aydogan C; Kutahya C; Allushi A; Yilmaz G; Yagci YBlock copolymer synthesis in one shot: concurrent metal-free ATRP and ROP processes under sunlight. Polym. Chem2017, 8, 2899–2903. [Google Scholar]

(183) Chen RY; Zhang WJ; Dai TZ; He JH; Ye HM; Wang YDFacile fabrication of water-soluble polyacrylic acid encapsulated core@shell upconversion nanoparticles via metal-free light induced surface initiated atom transfer radical polymerization. Mater. Lett2020, 273, 127874. [Google Scholar]

(184) Qian YQ; Han N; Gao XF; Gao XY; Li W; Zhang XXCellulose-based phase change fibres for thermal energy storage and management applications. Chem. Eng. J2021, 412, 128596. [Google Scholar]

(185) Cuneo T; Wang XF; Shi Y; Gao HFSynthesis of hyperbranched polymers via metal-free atrp in solution and microemulsion. Macromol. Chem. Phys2020, 221, 2000008. [Google Scholar]

(186) Niu TF; Jiang JY; Li SY; Ni BQ; Liu XM; Chen MQWell-defined high-molecular-weight polyacrylonitrile formation via visible-light-induced metal-free radical polymerization. Macromol. Chem. Phys2017, 218, 1700169. [Google Scholar]

(187) Wei DL; Xu YG; Liu CR; Zhai YX; Chen H; Bai LJ; Yang HW; Yang LX; Wang WX; Niu YZVisible light-induced metal-free atom transfer radical polymerization: An efficient approach to polyacrylonitrile. J. Polym. Sci., Part A: Polym. Chem2019, 57, 1265–1269. [Google Scholar]

(188) Ma L; Li N; Zhu J; Chen XDVisible light-induced metal free surface initiated atom transfer radical polymerization of methyl methacrylate on SBA-15. Polymers2017, 9, 58. [PMC free article] [PubMed] [Google Scholar]

(189) Xu YY; Li GC; Hu YF; Wang YBSynthesis of poly(N-isopropylacrylamide)-block-poly(tert-butyl methacrylate) block copolymer by visible light-induced metal-free atom transfer polymerization. Macromol. Chem. Phys2018, 219, 1800192. [Google Scholar]

(190) Sun Y; Li J; Zhao MY; Liu YT; Zhang JM; Lv CWPreparation of block copolymers via metal-free visiblelight-induced ATRP for the detection of lead ions. J. Appl. Polym. Sci2018, 135, 45863. [Google Scholar]

(191) Sun Y; Zhao MY; Liu YT; Fu LY; Li SY; Yang YFPreparation of erythromycin imprinted polymer by metal-free visible-light-induced ATRP and its application in sensor. J. Solid State Electrochem2019, 23, 583–590. [Google Scholar]

(192) Swisher NA; Corbin DA; Miyake GMSynthesis, characterization, and reactivity of N-alkyl phenoxazines in organo-catalyzed atom transfer radical polymerization. ACS Macro Lett. 2021, 10, 453–459. [PMC free article] [PubMed] [Google Scholar]

(193) Ma AY; Zhang JK; Wang N; Bai LJ; Chen H; Wang WX; Yang HW; Yang LX; Niu YZ; Wei DLSurface-initiated metal-free photoinduced ATRP of 4-vinylpyridine from SiO2 via visible light photocatalysis for self-healing hydrogels. Ind. Eng. Chem. Res2018, 57, 17417–17429. [Google Scholar]

(194) Bai LJ; Jiang XY; Sun ZX; Pei ZX; Ma AY; Wang WX; Chen H; Yang HW; Yang LX; Wei DLSelf-healing nanocomposite hydrogels based on modified cellulose nanocrystals by surface-initiated photoinduced electron transfer ATRP. Cellulose2019, 26, 5305–5319. [Google Scholar]

(195) Fan DC; Wang GL; Ma AY; Wang WX; Chen H; Bai LJ; Yang HW; Wei DL; Yang LXSurface engineering of porous carbon for self-healing nanocomposite hydrogels by mussel-inspired chemistry and PET-ATRP. ACS Appl. Mater. Interfaces2019, 11, 38126–38135. [PubMed] [Google Scholar]

(196) Jiang XY; Xi MZ; Bai LJ; Wang WX; Yang LX; Chen H; Niu YZ; Cui YM; Yang HW; Wei DLSurface-initiated PET-ATRP and mussel-inspired chemistry for surface engineering of MWCNTs and application in self-healing nanocomposite hydrogels. Mater. Sci. Eng., C2020, 109, 110553. [PubMed] [Google Scholar]

(197) Nguyen LT; Pham HQ; Nguyen DAS; Huynh KPH; Tran HL; Mai PT; Nguyen HT; Nguyen LTT; Truong TT10-(pyren-1-yl)-10h-phenothiazine and pyrene as organic catalysts for photoinitiated ATRP of 4-vinylpyridine. Polim.: Cienc. Tecnol2021, 31, e2021001. [Google Scholar]

(198) Singha NK; Kavitha A; Sarker P; Rimmer SCopper-mediated controlled radical ring-opening polymerization (RROP) of a vinylcycloalkane. Chem. Commun2008, 3049–3051. [PubMed] [Google Scholar]

(199) Lad J; Harrisson S; Mantovani G; Haddleton DMCopper mediated living radical polymerisation: interactions between monomer and catalyst. Dalton Trans. 2003, 4175–4180. [Google Scholar]

(200) Mittal A; Sivaram S; Baskaran DUnfavorable coordination of copper with methyl vinyl ketone in atom transfer radical polymerization. Macromolecules2006, 39, 5555–5558. [Google Scholar]

(201) Chen DF; Boyle BM; McCarthy BG; Lim CH; Miyake GMControlling polymer composition in organocatalyzed photoredox radical ring-opening polymerization of vinylcyclopropanes. J. Am. Chem. Soc2019, 141, 13268–13277. [PMC free article] [PubMed] [Google Scholar]

(202) Chen DF; Bernsten S; Miyake GMOrganocatalyzed photoredox radical ring-opening polymerization of functionalized vinylcyclopropanes. Macromolecules2020, 53, 8352–8359. [PMC free article] [PubMed] [Google Scholar]

(203) de Oliveira ERM; Vieira RPSynthesis and characterization of poly(limonene) by photoinduced controlled radical polymerization. J. Polym. Environ2020, 28, 2931–2938. [Google Scholar]

(204) Philipps K; Junkers T; Michels JJThe block copolymer shuffle in size exclusion chromatography: the intrinsic problem with using elugrams to determine chain extension success. Polym. Chem2021, 12, 2522–2531. [Google Scholar]

(205) Theriot JC; Miyake GM; Boyer CAN,N-Diaryl dihydrophenazines as photoredox catalysts for PET-RAFT and sequential PET-RAFT/O-ATRP. ACS Macro Lett. 2018, 7, 662–666. [PMC free article] [PubMed] [Google Scholar]

(206) Fu C; Xu F; Boyer CPhotoacid-mediated ring opening polymerization driven by visible light. Chem. Commun2016, 52, 7126–7129. [PubMed] [Google Scholar]

(207) Yilmaz GOne-pot synthesis of star copolymers by the combination of metal-free ATRP and ROP processes. Polymers2019, 11, 1577. [PMC free article] [PubMed] [Google Scholar]

(208) Discekici EH; St. Amant AH; Nguyen SN; Lee I-H; Hawker CJ; Read de Alaniz JEndo and exo diels-alder adducts: temperature-tunable building blocks for selective chemical functionalization. J. Am. Chem. Soc2018, 140, 5009–5013. [PMC free article] [PubMed] [Google Scholar]

(209) Hu X; Zhang YJ; Cui GP; Zhu N; Guo KPoly(vinylidene fluoride-co-chlorotrifluoroethylene) modification via organocatalyzed atom transfer radical polymerization. Macromol. Rapid Commun2017, 38, 1700399. [PubMed] [Google Scholar]

(210) Tan SB; Zhao YF; Zhang WW; Gao P; Zhu WW; Zhang ZCA light-mediated metal-free atom transfer radical chain transfer reaction for the controlled hydrogenation of poly(vinylidene fluoride-chlorotrifluoroethylene). Polym. Chem2018, 9, 221–227. [Google Scholar]

(211) Hu X; Li N; Heng TT; Fang L; Lu CHFunctionalization of PVDF-based copolymer via photo-induced p-anisaldehyde catalyzed atom transfer radical polymerization. React. Funct. Polym2020, 150, 104541. [Google Scholar]

(212) Lu CW; Wang CP; Yu J; Wang JF; Chu FXMetal-free ATRP “grafting from” technique for renewable cellulose graft copolymers. Green Chem. 2019, 21, 2759–2770. [Google Scholar]

(213) Lu CW; Guo XL; Wang CP; Wang JF; Chu FXIntegration of metal-free ATRP and Diels-Alder reaction toward sustainable and recyclable cellulose-based thermoset elastomers. Carbohydr. Polym2020, 242, 116404. [PubMed] [Google Scholar]

(214) Buss BL; Beck LR; Miyake GMSynthesis of star polymers using organocatalyzed atom transfer radical polymerization through a core-first approach. Polym. Chem2018, 9, 1658–1665. [PMC free article] [PubMed] [Google Scholar]

(215) Yin XZ; Wang LN; Zhang XM; Zhao HT; Cui Z; Fu P; Liu MY; Pang XC; Qiao XGSynthesis of amphiphilic star-shaped block copolymers through photo-induced metal free atom transfer radical polymerization. Eur. Polym. J2020, 126, 109557. [Google Scholar]

(216) Discekici EH; Pester CW; Treat NJ; Lawrence I; Mattson KM; Narupai B; Toumayan EP; Luo YD; McGrath AJ; Clark PG; et al. Simple benchtop approach to polymer brush nanostructures using visible-light-mediated metal-free atom transfer radical polymerization. ACS Macro Lett. 2016, 5, 258–262. [PubMed] [Google Scholar]

(217) Xu ZF; Tang B; Zhou H; Li H; Wang Y; Xu XZ; Zhou GFMetal-free and photo-controlled fabrication of fluorinated antifouling film. Polymer2020, 186, 122073. [Google Scholar]

(218) BelBruno JJMolecularly imprinted polymers. Chem. Rev2019, 119, 94–119. [PubMed] [Google Scholar]

(219) Wang N; Zhao XQ; Chen H; Bai LJ; Xu H; Wang WX; Yang HW; Wei DL; Yang LXFabrication of novel electrochemical immunosensor by mussel-inspired chemistry and surface-initiated PET-ATRP for the simultaneous detection of CEA and AFP. React. Funct. Polym2020, 154, 104632. [Google Scholar]

(220) Bai R; Sun Y; Zhao M; Han Z; Zhang J; Sun Y; Dong W; Li SPreparation of IgG imprinted polymers by metal-free visible-light-induced ATRP and its application in biosensor. Talanta2021, 226, 122160. [PMC free article] [PubMed] [Google Scholar]

(221) Yan JJ; Pan XC; Schmitt M; Wang ZY; Bockstaller MR; Matyjaszewski KEnhancing initiation efficiency in metal-free surface-initiated atom transfer radical polymerization (SI-ATRP). ACS Macro Lett. 2016, 5, 661–665. [PubMed] [Google Scholar]

(222) Xu L; Li HL; Wang LPpH-Sensitive, polymer functionalized, nonporous silica nanoparticles for quercetin controlled release. Polymers2019, 11, 2026. [PMC free article] [PubMed] [Google Scholar]

(223) Su HL; Xu L; Hu XJ; Chen FF; Li G; Yang ZK; Wang LP; Li HLPolymer grafted mesoporous SBA-15 material synthesized via metal-free ATRP as pH-sensitive drug carrier for quercetin. Eur. Polym. J2021, 148, 110354. [Google Scholar]

(224) Yan CN; Liu Q; Xu L; Bai LP; Wang LP; Li GPhotoinduced metal-free surface initiated ATRP from hollow spheres surface. Polymers2019, 11, 599. [PMC free article] [PubMed] [Google Scholar]

(225) Xu X; He JJ; Zeng YN; Yu CL; Zhang FControllable surface-initiated metal-free atom transfer radical polymerization of methyl methacrylate on mesoporous SBA-15 via reductive quenching. Eur. Polym. J2020, 131, 109724. [Google Scholar]

(226) Xu X; Zou Y; He JJ; Zeng YN; Yu CL; Zhang FAInsight into the effects of reaction conditions on metal-free surface-initiated atom-transfer radical polymerization of methyl methacrylate from SBA-15. J. Appl. Phys2020, 127, 115102. [Google Scholar]

(227) Huang L; Yu SX; Long W; Huang HY; Wen YQ; Deng FJ; Liu MY; Xu WY; Zhang XY; Wei YThe utilization of multifunctional organic dye with aggregation-induced emission feature to fabricate luminescent mesoporous silica nanoparticles based polymeric composites for controlled drug delivery. Micro-porous Mesoporous Mater. 2020, 308, 110520. [Google Scholar]

(228) Yang Y; Liu XG; Ye G; Zhu S; Wang Z; Huo XM; Matyjaszewski K; Lu YX; Chen JMetal-free photoinduced electron transfer-atom transfer radical polymerization integrated with bio-inspired polydopamine chemistry as a green strategy for surface engineering of magnetic nanoparticles. ACS Appl. Mater. Interfaces2017, 9, 13637–13646. [PubMed] [Google Scholar]

(229) Wang XB; You N; Lan FQ; Fu P; Cui Z; Pang XC; Liu MY; Zhao QXFacile synthesis of size-tunable superparamagnetic/polymeric core/shell nanoparticles by metal-free atom transfer radical polymerization at ambient temperature. RSC Adv. 2017, 7, 7789–7792. [Google Scholar]

(230) You N; Zhang CX; Liang YC; Zhang Q; Fu P; Liu MY; Zhao QX; Cui Z; Pang XCFacile fabrication of size-tunable core/shell ferroelectric/polymeric nanoparticles with tailorable dielectric properties via organocatalyzed atom transfer radical polymerization driven by visible light. Sci. Rep2019, 9, 1869. [PMC free article] [PubMed] [Google Scholar]

(231) Chen JY; Mao LC; Qi HX; Xu DZ; Huang HY; Liu MY; Wen YQ; Deng FJ; Zhang XY; Wei YPreparation of fluorescent cellulose nanocrystal polymer composites with thermo-responsiveness through light-induced ATRP. Cellulose2020, 27, 743–753. [Google Scholar]

(232) Zeng GJ; Liu MY; Heng CN; Huang Q; Mao LC; Huang HY; Hui JF; Deng FJ; Zhang XY; Wei YSurface polyPEGylation of Eu3+ doped luminescent hydroxyapatite nanorods through the combination of ligand exchange and metal free surface initiated atom transfer radical polymerization. Appl. Surf. Sci2017, 399, 499–505. [Google Scholar]

(233) Zeng GJ; Liu MY; Jiang RM; Heng CN; Huang Q; Mao LC; Hui JF; Deng FJ; Zhang XY; Wei YSurface grafting of Eu3+ doped luminescent hydroxyapatite nanomaterials through metal free light initiated atom transfer radical polymerization for theranostic applications. Mater. Sci. Eng., C2017, 77, 420–426. [PubMed] [Google Scholar]

(234) Jiang RM; Liu MY; Huang HY; Huang L; Huang Q; Wen YQ; Cao QY; Tian JW; Zhang XY; Wei YA novel self-catalyzed photoATRP strategy for preparation of fluorescent hydroxyapatite nanoparticles and their biological imaging. Appl. Surf. Sci2018, 434, 1129–1136. [Google Scholar]

(235) Garlets ZJ; Nguyen JD; Stephenson CRJThe development of visible-light photoredox catalysis in flow. Isr. J. Chem2014, 54, 351–360. [PMC free article] [PubMed] [Google Scholar]

(236) El Achi N; Bakkour Y; Adhami W; Molina J; Penhoat M; Azaroual N; Chausset-Boissarie L; Rolando CMetal-free ATRP catalyzed by visible light in continuous flow. Front. Chem2020, 8, 740. [PMC free article] [PubMed] [Google Scholar]

(237) Tan SB; Xiong J; Zhao YF; Liu JJ; Zhang ZCSynthesis of poly(vinylidene fluoride-cochlorotrifluoroethylene)-g-poly(methyl methacrylate) with low dielectric loss by photo-induced metal-free ATRP. J. Mater. Chem. C2018, 6, 4131–4139. [Google Scholar]

(238) Nikolaou V; Anastasaki A; Alsubaie F; Simula A; Fox DJ; Haddleton DMCopper(II) gluconate (a non-toxic food supplement/dietary aid) as a precursor catalyst for effective photo-induced living radical polymerisation of acrylates. Polym. Chem2015, 6, 3581–3585. [Google Scholar]

(239) Gilman H; Shirley DASome derivatives of phenothiazine. J. Am. Chem. Soc1944, 66, 888–893. [Google Scholar]

(240) Massie SPThe chemistry of phenothiazine. Chem. Rev1954, 54, 797–883. [Google Scholar]

(241) Shirley DA; Lehto EAThe metalation of 4-t-butyldiphenyl sulfone with n-butyllithium. J. Am. Chem. Soc1957, 79, 3481–3485. [Google Scholar]

(242) Dietrich LEP; Teal TK; Price-Whelan A; Newman DKRedox-active antibiotics control gene expression and community behavior in divergent bacteria. Science2008, 321, 1203–1206. [PMC free article] [PubMed] [Google Scholar]

(243) Price-Whelan A; Dietrich LEP; Newman DKRethinking ‘secondary’ metabolism: physiological roles for phenazine antibiotics. Nat. Chem. Biol2006, 2, 71–78. [PubMed] [Google Scholar]

(244) Palmer BD; Rewcastle GW; Atwell GJ; Baguley BC; Denny WAPotential antitumor agents. 54. Chromophore requirements for in vivo antitumor activity among the general class of linear tricyclic carboxamides