Switchable Polymerization Catalysts - Chemical Reviews (ACS

Nov 20, 2015 - Christopher W. Bielawski received a B.S. degree in chemistry from the University of Illinois, Urbana-Champaign, and a Ph.D. degree from...
0 downloads 12 Views 5MB Size
Review pubs.acs.org/CR

Switchable Polymerization Catalysts Aaron J. Teator,†,‡,∥ Dominika N. Lastovickova,†,‡,∥ and Christopher W. Bielawski*,‡,§ †

Department of Chemistry, The University of Texas at Austin, Austin, Texas 78712, United States Center for Multidimensional Carbon Materials (CMCM), Institute for Basic Science (IBS) and §Departments of Chemistry and Energy Engineering, Ulsan National Institute of Science and Technology (UNIST), Ulsan 689-798, Republic of Korea



Author Contributions Notes Biographies Acknowledgments Abbreviations References

1. INTRODUCTION The intrinsic physical properties and functionalities displayed by synthetic polymers are directly correlated to their respective microstructures. For example, complex biopolymers (e.g., proteins and DNA) are often responsible for governing various essential processes in living organisms and serve as inspirational illustrations of this central relationship. For this reason, the rational design and synthesis of polymeric materials with tailored microstructures of ever-increasing sophistication remains a central focus in macromolecular chemistry.1−5 Perhaps unsurprisingly, the degree of microstructural regularity and intricacy exhibited by the aforementioned biopolymers has yet to be attained in synthetic macromolecules. Such an achievement would represent the utmost control over a synthetic polymerization reaction and thus has been actively pursued. Since minor changes in tacticity, head-to-tail structure, sequence, and/or chain topology can have significant effects on the macroscopic properties of a synthetic polymer, the advent of controlled or living polymerization techniques provided the first means to transcend primary structural control and enabled access to a wealth of well-defined microstructures.6,7 Historically, the combination of judiciously timed monomer additions as well as careful attention to predetermined monomer reactivity ratios in conjunction with living polymerization methodologies has been utilized in attempts to manipulate the microstructures of polymeric materials. Although controlled monomer addition has produced various copolymers exhibiting relatively simple but well-defined microstructures, such an approach often becomes impractical as the desired level of polymer complexity increases. As such, many advanced polymeric architectures (e.g., symmetrically gradient, gradient-block, and other periodic copolymers) that are expected to possess novel physical properties are often synthetically intractable.8,9 An attractive solution to the aforementioned challenges may lie within the nascent field of switchable catalysis, whereby the chemical reactivity of a catalyst is selectively toggled between

CONTENTS 1. Introduction 1.1. Thermal Modulation 1.2. Chemical Modulation 1.3. Photochemical Modulation 1.4. Redox Modulation 1.5. Mechanical Modulation 2. Ring-Opening Metathesis Polymerization 2.1. Acid−Base Control 2.2. Mechanochemical Control 2.3. Redox Control 3. Ring-Opening Polymerization 3.1. Thermal Control 3.2. Photocontrol 3.3. Redox Control 3.4. Chemical Control 4. Copper-Mediated Reversible-Deactivation Radical Polymerization 4.1. Photocontrol 4.1.1. Ultraviolet Light Control 4.1.2. Visible Light Control 4.2. Electrochemical Control 5. Photoredox-Mediated Radical Polymerization 5.1. Transition Metal-Based Photoredox Catalysts 5.1.1. Ruthenium-Based Photoredox Catalysts 5.1.2. Iridium-Based Photoredox Catalysts 5.1.3. Niobium-Based Photoredox Catalysts 5.2. Organic-Based Photoredox Catalysts 6. Radical Polymerization via Controlled Bond Photolysis 6.1. Photocontrol 6.1.1. Visible Light Control 6.1.2. Ultraviolet Light Control 7. Alternative Polymerization Methods 7.1. Photocontrolled Olefin Polymerization 7.2. pH-Controlled Supramolecular Polymerization 8. Summary and Outlook Author Information Corresponding Author © 2015 American Chemical Society

1988 1988 1988 1988 1988 1989

1969 1970 1970 1970 1970 1970 1971 1971 1972 1972 1973 1973 1974 1975 1976 1977 1978 1978 1979 1979 1980 1980 1980 1981 1982 1983 1984 1984 1984 1986 1987 1987

Special Issue: Frontiers in Macromolecular and Supramolecular Science

1987 1987 1988 1988

Received: July 22, 2015 Published: November 20, 2015 1969

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

Figure 1. Utilizing external stimuli to switch between different forms of an active catalyst may allow for selective monomer discrimination among a mixture and result in polymeric materials with tailored architectures.

reactions: control over ligand dissociation to block coordination sites and modulation of electron density available at the metal center. While polymerizations that can be activated by changes in solution pH have been demonstrated, few reports exist where acid−base interactions have been utilized to switch the activity displayed by a catalyst over the course of a single polymerization reaction.

multiple, distinct states through the application of external stimuli.10−14 This has been a rapidly expanding area of research, with examples emerging that utilize a variety of different approaches to modulate the inherent reactivity of a catalytic system. The application of switchable catalysts to polymerization reactions has been a more recent development, but has nonetheless experienced an almost instantaneous proliferation. While early examples focused on the ability to cycle a polymerization reaction between on (active) and off (inactive) states, relatively advanced switching characteristics that allow for selective monomer discrimination are now being explored and may lead to new classes of polymeric materials that feature sophisticated microstructures (Figure 1). Modulating the activity and/or selectivity of a polymerization catalyst can be achieved through a number of different methodologies. These methods often entail the incorporation of stimulus-responsive functional groups into known catalytic systems as well as the direct alteration of the intrinsic properties (i.e., oxidation state) of the catalytically active species. Ideally, a chosen stimulus must transmit energy in a discriminate manner to allow for site-specific transformations. Therefore, a careful consideration of the available stimuli (e.g., photo, thermal, redox, etc.) is essential to achieve high degrees of spatial and temporal control.

1.3. Photochemical Modulation

Thus far, electromagnetic radiation has been the most widely studied stimulus. Light is particularly attractive as a stimulus because it is noninvasive and offers an almost unparalleled level of control through careful selection of irradiation wavelength and power. The wide variety of available photosensitizers combined with the general ease in which UV and/or visible light can be introduced has enabled a surge in the number of photoswitchable polymerization catalysts. Additionally, the establishment of photoredox catalysis as an efficient means to convert light energy into chemical energy has resulted in new polymerization methods with precise control over the catalyst and/or polymer oxidation state.18−20 1.4. Redox Modulation

As many metal complexes exhibit relatively low redox potentials, external control over oxidation and reduction events can impart significant control over catalytic processes. Historically, redox control has been accomplished electrochemically as well as via the introduction of chemical oxidants or reductants and has given rise to redox switchable polymerizations. More recently, photoredox catalysts, which utilize excited state electron transfer processes, have provided an alternative means to modulate oxidation states and have helped to overcome some of the limitations (e.g., dependencies on certain electrodes or electrolytes) that may hamper other methods.

1.1. Thermal Modulation

Perhaps the simplest method to induce a chemical transformation is through thermal activation. However, despite the relative ease with which temperature can be adjusted in the laboratory, the application of thermal energy as a reactivity switch is a fundamentally more difficult task. Elevated reaction temperatures result in a broad distribution of energies, which can often give rise to undesirable side reactions and/or decomposition pathways. Indeed, while changes in temperature have been exploited to activate latent polymerization catalysts,14−17 the application of heat to switch polymerization reactions is relatively rare.

1.5. Mechanical Modulation

In contrast to the above-described methods, the integration of mechanoresponsive functionalities into predesigned molecular scaffolds has emerged as a means to direct chemical reactivity.21,22 The incorporation of polymer chains into catalytically active metal complexes provides handles to which extrinsic forces can be applied, subsequently inducing mechanochemical transformations. Ultrasonication has proven to be an efficient method for the application of external force as ultrasound-induced cavitation events produce substantial solvodynamic shear that is maximized in the middle of polymer chains and can result in site-specific bond activations. Although

1.2. Chemical Modulation

The introduction of chemical stimuli regularly elicits structural changes in biomolecules and therefore represents an attractive trigger to mimic the control seen in natural processes. In fact, the incorporation of pH-responsive ligands into transition metalbased catalysts was one of the earliest forays into the development of switchable polymerization systems. Conceptually, there have been two fundamental approaches to utilizing acid−base chemistry to affect the outcome of polymerization 1970

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

Scheme 1. Reversible Inhibition of the ROMP of Cyclooctene via the Addition of a Competing MIM Ligand

Scheme 2. Synthesis of Poly(norbornene) with Tunable E/Z Ratios through the Addition of Acid to a pH-Sensitive ROMP Catalyst

2.1. Acid−Base Control

the scope of mechanochemically switchable polymerization systems continues to develop, this method stands out for harnessing acoustic energy to selectively trigger useful chemical transformations. In this review we aim to highlight advances in the burgeoning field of switchable polymerization catalysts and include examples that range from ring-opening metathesis polymerizations and ring-opening polymerizations to controlled radical polymerizations and beyond. Focus will be directed toward catalysts that demonstrate truly switchable behavior, that is, chemical reactivity that can be selectively modulated at will over the course of a single reaction. Polymerizations that are stimulus-induced or otherwise regulated but are not necessarily switchable by this definition fall outside the scope of this review and have been elegantly covered elsewhere.23−27 Likewise, switchable reagents28 that efficiently modulate polymerization activity but operate through the use of multiple reaction vessels and/or require a significant change in reaction conditions as part of the switching process will not be discussed. In the following sections, various thermally, chemically, photochemically, and other stimulus controlled “switches” will be described within the context of the polymerizations to which they have been applied.

In a pioneering example, P’Pool and Schanz disclosed the first reversible complete inhibition of a ROMP reaction.31 The addition of excess N-methylimidazole (MIM) to the Grubbs firstgeneration catalyst (1) resulted in the formation of a new ruthenium complex (2) (Scheme 1). The substitution of one tricyclohexylphosphine (PCy3) ligand with two MIM ligands was reported to have an immediate effect on catalytic activity in the ROMP of cyclooctene as the resulting complex was found to be metathesis inactive. No monomer conversion was observed after 24 h, at which point the catalyst could be reactivated through the addition of excess H3PO4. Protonation and subsequent liberation of the MIM ligands released the active monophosphine complex 3. The ROMP reaction proceeded virtually instantaneously at an accelerated rate relative to that observed with catalyst 1 and proceeded to a high monomer conversion. While the in situ generation of 2 caused complete inhibition of the ROMP reaction, the complex proved to be a competent, albeit slow, ROMP catalyst when isolated. Thus, the inactivity of complex 2 can be attributed to the formation of a catalyst with limited metathesis activity as well as the presence of numerous donor ligands, including liberated PCy3, that may suppress the rate of the ROMP reaction. Similarly, the observed increase in monomer conversion relative to that observed with 1 was likely a result of the presence of highly metathesis-active complex 3 as well as the protonation of the remaining free donor ligands. Another pH-responsive ruthenium-based ROMP catalyst was described by Plenio and co-workers, who demonstrated switchable stereocontrol in a ROMP reaction by modulating the electronic properties of a Hoveyda−Grubbs-type metathesis catalyst.32 The incorporation of secondary amines into the aryl substituents of an N-heterocyclic carbene (NHC) ligand produced the pH-responsive catalyst 4 (Scheme 2). Complex 4 initiated the ROMP of norbornene and furnished a polymer that exhibited a polydispersity index (Đ) of 1.65 and an E/Z ratio of 0.78 in nearly quantitative yield. Exposure of 4 to acidic conditions resulted in the rapid protonation of the amine functional groups to generate 5. The ammonium-containing complex 5 promoted the formation of poly(norbornene) with an

2. RING-OPENING METATHESIS POLYMERIZATION Ring-opening metathesis polymerization (ROMP) is a powerful technique for the preparation of highly regular polymeric materials from cyclic olefins and has given rise to a number of unique polymer architectures.29,30 The diverse range of commercially available, highly active, and functional group tolerant ROMP catalysts has resulted in an abundance of reports that provide valuable insight into monomer scope, stereochemical outcome, and reaction mechanism. Although the development of switchable ROMP catalysts has only recently been pursued, significant progress has already been made. Numerous studies have reported that the activity of ROMP catalysts can be selectively turned between on (i.e., relatively active) and off (i.e., relatively inactive) states upon the application of mechanical force or via the manipulation of an oxidation state or pH. 1971

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

Scheme 3. Mechanochemical Generation of a ROMP-Active Ru Complex

2.3. Redox Control

E/Z ratio of 1.04 and a slightly lower molecular weight distribution (Đ = 1.55), effectively switching the stereochemical outcome of the ROMP reaction. The E/Z ratio of the polymeric product was further tuned by introducing acid at various times over the course of a polymerization reaction. Comparative studies with an analogous Grubbs thirdgeneration-type catalyst indicated that the structure of the precatalyst had no significant effect on the E/Z ratios of the resulting poly(norbornene). Since protonation can be assumed to have a minimal effect on the steric environment about the metal center, the authors postulated that electronics were responsible for the aforementioned stereochemical switch. Substitution of the NEt2 groups with less donating substituents resulted in poly(norbornene)s with E/Z ratios that were lower than those obtained with catalyst 4, which was opposite the effect observed upon generation of protonated catalyst 5. Similarly, Schanz and co-workers previously reported that a reduction in ROMP propagation rates upon the protonation of similar complexes could be the result of increased electron density at the metal center.33 DFT-derived Mulliken charge calculations offered some insight into the observed stereochemical outcomes and indicated a decrease in positive charge at the metal center due to the reduced π-acceptor capability of the protonated amine-functionalized NHC ligand.33

In 2013, Plenio and co-workers reported an oxidation-triggered ROMP by incorporating a latent ferrocene chelate into a Grubbstype catalyst.36 A ferrocene-appended Schiff base ligand displayed tunable donicity upon oxidation of the Fe center, which promoted ligand dissociation and effectively activated the catalyst toward the ROMP of cyclooctene. Later in 2013, Bielawski and co-workers reported a ruthenium−indenylidene metathesis-active complex 8, which featured a redox-active diaminocarbene−[3]ferrocenophane (FcDAC) ligand in place of the prototypical NHC (Scheme 4).37 Oxidation of the FeII Scheme 4. Remote Tuning of ROMP Activity via Redox Modulation of a Ferrocenophane Ligand

center in complex 8 was found to occur at a lower potential relative to that of the RuII center, therefore allowing site-selective redox activity and controlled generation of complex 9. The addition of 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) to an ongoing ROMP reaction of cis,cis-1,5-cyclooctadiene (COD) resulted in attenuation of the observed polymerization rate constants (kred/kox = 37.5). Subsequent addition of decamethylferrocene (Fc*) reverted the FeIII center to FeII and restored approximately 35% of the initial catalytic activity. The authors postulated that the reduced donating ability of the oxidized FcDAC ligand was responsible for the diminished ROMP activity, although the possibility of concurrent oxidation of the Ru center could not be excluded. Building on these results, the same group reported switchable ROMP activity through the oxidation and reduction of the commercially available Grubbs second-generation catalyst.38,39 Precipitation of the catalytically active complex was observed upon oxidation of the RuII center to RuIII, indicating that Rucentered redox processes could also be used to modulate catalytic activity. Similar to the method described above, the addition of DDQ to an ongoing ROMP reaction of COD significantly decreased the observed reaction rate constants (kred/kox = 80). Subsequent addition of Fc* to the reaction mixture restored roughly 27% of the initial catalytic activity, and the polymerization proceeded to high conversion. The off/on behavior of this system was attributed to the reversible solubility of the active catalyst as the addition of DDQ resulted in the precipitation of the oxidized catalyst, and therefore could also be used to remove the catalyst from the reaction mixture.

2.2. Mechanochemical Control

Sijbesma and co-workers reported the synthesis of a metathesisactive ruthenium complex (6) bearing two polymer-appended NHC ligands and demonstrated switchable activity in the ROMP of cyclooctene (Scheme 3).34 Subjecting complex 6 to ultrasonication forced the dissociation of one of the NHC ligands and released the highly active catalyst 7. In the absence of sonication, precatalyst 6 was effectively in an inactive state as minimal conversion of monomer to polymer was observed after 2 h. Acoustic activation initiated the aforementioned ROMP reaction, and the authors reported that further conversion ceased each time the sonication was interrupted. The polymerization proceeded to high overall conversion and produced a relatively high molecular weight (Mn = 40 kDa) poly(octenamer). The polymer’s polydispersity index was slightly broad (Đ = 1.6) and may reflect a slow, but continuous, mechanochemical liberation process to form the active catalyst. Regardless, the slow initiation likely facilitated the switchable feature of the polymerization as NMR data supported the decomposition of 7 during periods without sonication, which impeded the ROMP reaction. Subsequent activation of the remaining precatalyst 6 promoted the ROMP of cyclooctene and gave rise to the observed off/on behavior.35 Indeed, this method validated ultrasound-induced mechanochemical activation as a useful switch for controlling chemical reactivity. 1972

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

Scheme 5. Mechanism of Reversible ROMP Activation and Deactivation via Photoirradiation of a Pyrylium Organocatalyst

3.1. Thermal Control

Recently, Boydston and co-workers reported an elegant metalfree, switchable ROMP by employing an organic photoredox catalyst (Scheme 5).40 Visible light (λ = 450−480 nm) irradiation of a pyrylium photocatalyst generated an excited species capable of oxidizing electron-rich vinyl ethers. The resulting vinyl ether cation reacted with norbornene to form a [2 + 2] adduct, followed by rapid ring-opening to initiate a ROMP reaction. A subsequent single-electron transfer event allowed for reversible deactivation through reduction of the propagating radical chain ends and regeneration of the original pyrylium photocatalyst. While no polymerization was observed in the absence of light, further visible light irradiation enabled reactivation of the polymer chain growth. The authors were able to achieve excellent temporal kinetic control through intermittent light exposure, effectively shutting down the ROMP propagation multiple times over the course of a single reaction. In addition, this method effectively demonstrated that externally controlled ROMP reactions are not restricted to transition metal complexes.

Thermally switchable organocatalytic ROPs utilizing NHCs were first examined by Hedrick, Waymouth, and co-workers in 2005.45 As shown in Scheme 6, triazolylidene 10 was demonstrated to Scheme 6. Thermally Reversible Organocatalytic ROP That Uses a Triazolylidene−Alcohol Adduct as the Dormant Form of the Active NHC Catalyst

3. RING-OPENING POLYMERIZATION The ring-opening polymerization (ROP) of cyclic esters serves as a resourceful and versatile method for the production of poly(ester)s, many of which provide eco-friendly alternatives to petroleum-based poly(olefin)s. The cross-discipline interest to produce well-defined poly(ester)s has resulted in a large number of available methods for performing ROP reactions that utilize organocatalysis as well as transition metal catalysis.41−43 Given the utility of these ROP-prepared materials, it is desirable to possess full control over the corresponding polymerization process. While remarkable progress has been made, access to poly(ester)s with rationally designed microstructures, and thus tailored physical properties, remains highly sought-after and has driven the quest for switchable ROP catalysts.44 External stimuli, such as light and heat, as well as the chemical alteration of ROP catalysts to promote a change in the electronic properties of the catalytic centers, have been used to selectively induce and inhibit the ROP of various monomers.

catalyze the ROP of cyclic esters (i.e., L-lactide and βbutyrolactone46) in the presence of an alcohol initiator. While the triazolylidene reacted with the alcohol to produce the ROPinactive adduct 11, subsequent heating to 90 °C reversed the reaction. The free carbene 10 initiated the ROP reaction by promoting the ring-opening of various cyclic esters to form the zwitterionic complex 12. After protonation by the alcohol initiator, the growing polymer chain was released and the free carbene 10 was regenerated through nucleophilic attack by the counterion. Subsequent cooling of the reaction mixture to 20 °C resulted in alcohol addition to the carbene center and yielded the dormant NHC adduct 11, hindering further propagation. Heating to 90 °C regenerated the active catalyst and restored polymer growth. Both L-lactide (L-LA) and β-butyrolactone 1973

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

underwent ROP when heated to 90 °C in the presence of an alcohol initiator and resulted in well-defined polymers with narrow dispersities (Đ = 1.08−1.29). Furthermore, the ROP of LLA could be repeatedly inhibited and restarted by cycling the reaction temperature between 20 and 90 °C, respectively. Collectively, the reversibility between the dormant NHC− alcohol adduct and the active free carbene afforded excellent control over the respective ROP reactions.

that regulated sequence-controlled polymerizations of biological structures, this photoswitchable ROP process employed αcyclodextrin (α-CD) as the catalytic host. The cinnamoylfunctionalized α-CD 2-O-trans-cinnamoyl-α-cyclodextrin (2trans-CIO-α-CD) was demonstrated to function as a ROP-active catalyst and facilitated the polymerization of δ-valerolactone (VL) at 100 °C. Upon entry of the monomer guest into the cavity of the host, the carbonyl of the VL was activated toward nucleophilic attack and facilitated polymer chain growth. Under these conditions, the ROP of VL proceeded with excellent monomer conversion (82%) and afforded a polymer with a relatively high degree of polymerization (DP = 41.2). UV irradiation (λ = 280 nm) of the reaction mixture was found to convert 2-trans-CIO-α-CD to its cis isomer (2-cis-CIO-α-CD), which increased the steric bulk around the catalytic cavity and effectively prevented monomer incorporation. As a result, the monomer conversion was relatively low (12%) and polymers with a relatively low degree of polymerization (DP = 8.1) were produced. In addition to photoisomerization, photocyclization was another light-controlled process that was demonstrated to regulate the ROP of various monomers. In 2013, Neilson and Bielawski developed a photoswitchable organocatalytic polymerization of cyclic lactones based on an in situ generated photochromic NHC.48 As shown in Scheme 8, the ROP-active dithienylethene-annulated NHC 14 was prepared by the addition of sodium hexamethyldisilazide (NaHMDS) to the corresponding precatalyst 13. In the presence of an alcohol initiator (i.e., benzyl alcohol), photochromic NHC 14 effectively catalyzed the ROP of VL as well as ε-caprolactone (CL), likely via the formation of imidazolium alkoxide 15. NMR data indicated that 15, which represented the resting state of the catalyst, underwent photoinduced electrocyclic ring closure to generate the catalytically inactive covalently bound NHC−alcohol adduct 16 upon exposure to UV radiation (λ = 313 nm). The extended conjugation along the backbone of photocyclized 16 resulted in a more electron-deficient carbenoid center that prevented the release of the alkoxide unit and rendered 16 relatively unreactive

3.2. Photocontrol

Photoinduced transformations that modify the steric environment about a catalytic center as well as those that lead to a change in the electronic properties of a catalyst have both been implemented to modulate ROP activity. Harada and co-workers designed a switchable ROP system that took advantage of the trans/cis photoisomerization of a cinnamoyl group to reversibly change the steric hindrance around the active site of a catalyst, resulting in significant variations in the respective polymerization rates (Scheme 7).47 On the basis of a host−guest enzyme model Scheme 7. Photoswitchable Polymerization of VL Controlled via trans/cis Isomerization of 2-CIO-α-CD That Reversibly Prevents Monomer Incorporation into the Catalytically Active Cavity

Scheme 8. Photoswitchable ROP Activity Using a Photoinduced Electrocyclic Ring Closure of a Photochromic DiaryletheneAnnulated NHC Organocatalyst

1974

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

Scheme 9. Reversible Inhibition of ROP Activity by Modifying the Lewis Acidity of a Ferrocene-Appended Ti−Salen Complex

Scheme 10. Y- and In-Based Redox Switchable Polymerizations of L-LA and Trimethylene Carbonate, Respectively

carbonate when the ferrocene moiety was oxidized to cationic 20 using ferrocenium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (FcBArF). Subsequent reduction of the FeIII center to FeII using cobaltocene (CoCp2) decreased the electron withdrawing ability of the phosfen ligand and enhanced polymer chain growth. The ROP of L-LA could be switched between on (i.e., relatively active) and off (i.e., relatively inactive) states numerous times via the oscillation between the two different oxidation states of the ferrocene component of the catalyst. Even after multiple switching cycles, the YIII catalyst 20 did not appear to suffer a loss in activity, as the polymerization rate constant from before and after the oxidation cycle remained relatively constant, and afforded well-defined polymers with narrow dispersities (Đ = 1.03−1.07). The opposite effect was observed when In-based catalyst 21 was used to facilitate the ROP of trimethylene carbonate: in its oxidized state, the FeIII-containing cationic complex 22 initiated the polymer growth but appeared to be inhibited in its reduced (neutral) FeII state. Diaconescu and co-workers also examined a series of zirconium- and titanium-based catalysts comprising ferrocenyl substituents that were capable of selective discrimination between two different monomers upon oxidation or reduction (Scheme 11).51 In the reduced (neutral) state, the ZrIV and TiIV complexes promoted the polymerization of L-LA at 100 °C but were inactive toward the ROP of CL. Oxidation of the ROP catalysts with FcBArF to form the cationic FeIII-containing derivatives resulted in the opposite reactivity where the oxidized catalysts induced the ROP of CL yet resulted in no significant monomer consumption of L-LA. Furthermore, this discrimination enabled a one-pot synthesis of block copolymers of L-LA and CL. Initially, the neutral TiIV complex 23 was used to convert L-LA to polymer in 58% conversion. Subsequent formation of the cationic derivative 24 via the addition of FcBArF facilitated the polymerization of CL (18% conversion) and ultimately afforded poly(L-LA-b-CL) as a well-defined (Đ = 1.12) copolymer. The ability to select between different monomers from within a mixture appears to be exclusive to the aforementioned TiIV complex 23 as similar selectivities were not observed with the ZrIV-based catalyst 25. After the initial polymerization of L-LA,

toward ROP (kamb/kUV = 114). Visible light irradiation (λ > 500 nm) reversed the photocyclization, regenerated the catalytically active imidazolium alkoxide 15, and restored the ROP activity (kvis/kUV = 17). While the reversible photocyclization of the NHC organocatalyst allowed for the ROP reaction to be switched between off (i.e., relatively inactive) and on (i.e., relatively active) states upon selective irradiation, each switching cycle resulted in approximately 13% catalyst decomposition. 3.3. Redox Control

Redox-induced changes in the electronic properties of the catalytic centers in transition metal catalysts have also been shown to affect the rates of ROP reactions. The first instance of redox-mediated ROP activity was described in 2006 by Gibson and co-workers, who incorporated a ferrocenyl-substituted salen ligand into a titanium-based polymerization catalyst to generate 17 (Scheme 9).49 When the ferrocene components of 17 existed in their reduced FeII form, the complex catalyzed the ROP of raclactide (rac-LA) at 70 °C, resulting in 18% monomer conversion after 8 h. Oxidation of the ferrocene moieties by silver triflate (AgOTf) formed the positively charged derivative 18. The decreased electron density at the TiIV catalytic site resulted in more tightly bound alkoxy ligands, which hindered the growth of polymer chains (kred/kox ≈ 30). Subsequent introduction of ferrocene (FeCp2) to the reaction mixture reduced the FeIII centers to their neutral FeII states and restored the catalytic activity of 17. Characterization of the polymers produced indicated that their molecular weight distributions were relatively low (Đ < 1.2). Likewise, the polymerizations catalyzed by 17 remained well-controlled even after multiple redox switching events. Another remotely controlled redox system was reported by Diaconescu and co-workers, who achieved excellent regulation of the polymerization rates of various monomers using either yttrium- or indium-based alkoxide catalyst 19 or 21, respectively (Scheme 10).50 The yttrium- and indium-based catalysts contained ferrocene-appended phosfen Schiff base ligands, which acted as remote redox switches. The catalytic activity displayed by the YIII complex 19 was found to be relatively inactive toward the polymerization of L-LA or trimethylene 1975

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

center to CeIV with FcBArF halted the polymer growth, the subsequent addition of CoCp2 regenerated 29 and resumed the ROP of meso-LA, which yielded polydisperse polymers (Đ = 2.25) in high conversion. Recently, iron-based complexes have also been utilized in redox switchable ROP systems. Byers and co-workers reported a bis(imino)pyridine−FeII catalyst that effectively polymerized rac-LA in the presence of an alcohol initiator.54 Upon the addition of FcPF6, the rate of the polymerization reaction was diminished due to the formation of the ROP-inactive FeIII analogue. Subsequent addition of a reductant (i.e., CoCp2) restored the initial FeII catalyst, which again facilitated the polymerization of rac-LA. The in situ redox switch could be performed multiple times over the course of a single ROP of racLA and proceeded with excellent control to yield well-defined polymers with narrow molecular weight distributions (Đ = 1.058−1.199). The opposite reactivity pattern was observed during the redoxcontrolled polymerization of CL using an alcohol-activated FeIIICl3 catalyst with a coordinating pincer ligand. Lang and coworkers reported that the reduced FeII form of the catalyst was inactive toward the polymerization of CL, even in the presence of excess alcohol initiator at 100 °C.55 Subsequent addition of FcPF6 formed the corresponding oxidized FeIII complex, which underwent protonolysis with the alcohol initiator to initiate the ROP of CL. The addition of CoCp2 effectively shut down the polymerization process as the additional consumption of the remaining monomer was not observed. The polymerization of CL using this system was highly selective, yielding well-defined polymers with narrow dispersities (Đ = 1.08−1.18).

Scheme 11. Ti- and Zr-Based ROP Catalysts Containing Redox Switchable Ligands That Enable the Preferential Polymerization of Certain Monomers

mediated by the reduced derivative of 25, the addition of FcBArF as an oxidant did not promote the ROP of CL. The authors ascribed the inability of the corresponding in situ oxidized ZrIV complex 26 to induce the polymerization of CL to the high Lewis acidity of the Zr center. In addition, the strong coordination of LLA to the active site blocked the complexation of CL, and thus prevented polymerization. In 2011, the same group demonstrated that redox switchable ROP systems were not limited to remote redox-sensitive ligands and could be extended to the direct modulation of the oxidation state of the catalytically active metal center.52 As shown in Scheme 12, the salen-ligated CeIII complex 27 induced the ROP of L-LA as well as rac-LA and resulted in poly(LA)s with narrow dispersities (Đ = 1.12−1.34). The subsequent addition of FcBArF effectively oxidized the CeIII center to a CeIV derivative and resulted in the formation of the charged complex 28, which was found to be inactive toward the polymerization of L-LA. The addition of CoCp2 reversed the oxidation reaction and produced the initial CeIII complex 27, which in turn restored the polymerization of L-LA. The redox switching was repeated multiple times without significant losses in conversion or polymerization rate, although slightly broadened molecular weight distributions (Đ = 1.53−1.73) were observed. In 2013, Okuda and co-workers reported that the polymerization of meso-lactide (meso-LA) could also be controlled with a redox switchable Ce-based catalyst (Scheme 13).53 Two OSSOtype bis(phenolate) ligands were incorporated into a neutral CeIV complex (29), which could be reduced to generate the anionic CeIII complex 30. The complex 30 was found to catalyze the polymerization of meso-LA at 40 °C, with 20% monomer conversion achieved after 45 min. While oxidation of the CeIII

3.4. Chemical Control

The addition and removal of ligands to reversibly block the active sites of ROP catalysts has also been used to control polymerization reactions. In 2010, Mirkin and co-workers constructed the triple-layer catalyst 31, which featured a ROP-active AlIII center coordinated to two phosphinoamine-bearing RhICl complexes via a functionalized salen ligand (Scheme 14).56 In the presence of chloride anions, which acted as allosteric effector agents, the AlIII catalytic center was sterically accessible toward monomer coordination and facilitated the ROP of CL to yield well-defined polymers (Đ = 1.17−1.18). The chloride anions were removed from the RhI centers upon the introduction of NaBArF, which enabled the chelation of the phosphinoamine ligands. The resulting complex 32 featured enhanced steric hindrance around the AlIII center that concealed the active site and effectively stopped the growth of polymer chains. The blockage of the catalytic center was selectively reversed upon addition of acetonitrile to re-expose the active site, enable CL coordination, and restore the ROP activity.

Scheme 12. Redox Switchable ROP of L-LA Based on the Tunable Oxidation State of a Ce−Salen Catalyst

1976

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

Scheme 13. Redox Switchable ROP of meso-LA Using a Ce−OSSO Complex

Scheme 14. Allosteric Control over a ROP-Active Al Center To Allow Reversible Inhibition of CL Polymerization

Scheme 15. Regulation of the Polymerization of CL via Reversible Base-Facilitated CO2 Capture

chain into a dormant state (34). The polymerization activity could be recovered by purging the reaction mixture with nitrogen gas for approximately 5−10 min at ambient temperature or 2−3 min at 40 °C to detach the carbonate end groups. The organocatalyzed polymerization of CL was successfully turned off and on (i.e., between relatively inactive and active states, respectively) multiple times without significant losses in the polymerization rates between the switching cycles and afforded relatively well-defined polymers (Đ = 1.2−1.3). Although increasing the [M]/[I] ratio to 200/1 appeared to hamper the switchable activity of the system, the method operated at relatively low [M]/[I] ratios (e.g., [M]/[I] = 50/1) and demonstrated that the reversible addition and removal of small molecules can selectively promote or inhibit the growth of polymer chains.

4. COPPER-MEDIATED REVERSIBLE-DEACTIVATION RADICAL POLYMERIZATION Since the first recognition of the “living” radical polymerization of styrene in 1957,58 substantial efforts have been directed toward the development of highly controlled radical polymerizations (CRPs) to afford well-defined polymers. In particular, significant progress has been made in modulating coppermediated radical polymerizations, which are typically induced by a CuI activator complex that reacts with an alkyl halide to generate a dormant CuII complex and an active radical species that facilitates polymerization (Scheme 16).59 The polymerization process can be deactivated by CuII species, which quenches the propagating radicals and, ultimately, reduces the polydispersities of the polymers produced. Since coppermediated radical polymerizations depend on the equilibrium between the activating CuI complex and the deactivating CuII complex, it is desirable to efficiently control their precise ratio in the reaction mixture and selectively switch between these two oxidation states to enable on-demand tuning of the polymerization process. However, despite the numerous techniques to generate reduced CuI species, only a limited number of stimuli

In addition to the use of allosteric effectors to selectively expose and sterically shield the active site of a ROP catalyst, the addition and removal of competing coordinating agents was also demonstrated to be an effective method for chemically switching the rates of polymerization reactions. Coulembier and coworkers reported an organocatalytic ROP system whose activity toward the polymerization of CL and trimethylene carbonate could be inhibited in the presence of carbon dioxide.57 As illustrated in Scheme 15, a 10:1 mixture of 1,5,7-triazabicyclododecene (TBD) and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) induced a ROP in the presence of an alcohol initiator. The introduction of CO2 resulted in its reversible fixation by the alcohol terminus of the growing polymer chain 33 and concomitant protonation of DBU, thus forcing the propagating 1977

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

the CuII species, producing hydroxymethyl radicals and formaldehyde, respectively. The additional formation of CuII limited the concentration of propagating radicals and therefore resulted in poly(MMA)s with narrow dispersities (Đ = 1.06−1.13). Moreover, the reaction could be switched between off (i.e., relatively inactive) and on (i.e., relatively active) states by periodically varying the light source. The same group later demonstrated the utility of the selectively controlled on and off polymerization processes in the synthesis of block copolymers of MMA and azide-functionalized styrene.62,63 By modifying the reaction conditions and choosing the proper ligand, initiator, and solvent, photoinduced ATRP and CuAAC were simultaneously used to grow polymers in a well-controlled manner. Matyjaszewski, Yagci, and co-workers further expanded the versatility of UV-controlled radical polymerizations by altering the identity of the photoinitiator under various conditions.64 The authors applied UV switchable control to the polymerization of oligo(ethylene glycol) monomethyl ether methacrylate (OEOMA) in inverse microemulsion media. The presence of different surfactants (i.e., poly(oxyethylene) (3) oleyl ether and poly(oxyethylene) (6) oleyl ether) in hexane along with aqueous monomer solutions containing OEOMA resulted in the formation of biphasic reaction mixtures. The aqueous layer also contained the Ciba Irgacure 2959 (1-[4-(2-hydroxyethoxy)phenyl]-2-hydroxy-2-methyl-1-propan-1-one) photoinitiator, a poly(ethylene glycol) bromide macroinitiator, and a CuIIBr2 catalyst ligated to PMDETA or tris(2-pyridylmethyl)amine (TPMA) and produced micelles that encapsulated the watersoluble reactants upon stirring. UV irradiation (λ = 350 nm) stimulated the Irgacure photoinitiator to reduce the CuII complex to a CuI species, which then facilitated the polymerization process via radical formation from the macroinitiator and yielded relatively well-defined (Đ = 1.20−1.40) polymers. The polymer chain growth was hindered in the absence of UV light, but could be restored upon subsequent photoirradiation. The Irgacure photoinitiator acted as an effective switch to modulate the polymerization kinetics, while the polymer molecular weights were determined by the size of the micelles. Irgacure was not the only photoinitiator to successfully reduce the CuII deactivator upon exposure to UV radiation. In 2014, Yagci and co-workers reported the use of zinc oxide nanoparticles and mesoporous graphitic carbon nitride as effective photoinitiators of copper-mediated polymerizations.65,66 In the presence of an ethyl α-bromoisobutyrate (EBiB) initiator, a CuIIBr2/PMDETA complex facilitated the polymerization of MMA upon exposure to UV radiation (λ = 350 nm). However, in the absence of UV light, CuI species were not being regenerated from their corresponding CuII precursors, which effectively prevented polymer growth. The polymerization process could be later restarted upon exposure to UV radiation. The methods involving zinc oxide nanoparticles as well as the mesoporous graphitic carbon nitride exhibited high control over the corresponding polymerization reactions and afforded materials with low polydispersities (Đ = 1.11−1.28).

Scheme 16. General Mechanism for Switchable Cu-Mediated Reversible-Deactivation Radical Polymerizations

have been shown to facilitate the in situ modulation of coppermediated radical polymerizations. 4.1. Photocontrol

4.1.1. Ultraviolet Light Control. In 2010, Kwak and Matyjaszewski combined an iniferter polymerization with atom transfer radical polymerization (ATRP) to selectively regulate the rates of radical polymerizations (Scheme 17).60 While iniferters typically serve as the initiators, transfer agents, and terminating agents of various polymerization processes, the iniferter 2-(N,N-diethyldithiocarbamyl)isobutyric acid ethyl ester (EMADC) was used as a photocleavable moiety to modulate the polymerization rate of methyl methacrylate (MMA). Exposure of a mixture of EMADC, CuIIBr2 catalyst, and hexamethyltriethylenetetramine (HMTETA) ligand to UV radiation (λ = 230−400 nm) at 30 °C facilitated the polymerization of MMA and produced well-defined (Đ = 1.27) polymers. In the absence of UV radiation, the formation of the dithiocarbamate radicals that promoted the polymerization process was suspended and the rate of polymerization decreased significantly (kUV/kamb ≈ 67). The polymer chain growth of MMA could be restored upon re-exposure to UV radiation, and the photoswitchable process was repeated multiple times. The presence of CuIIBr2 and HMTETA ligand was required to ensure efficient chain transfer reactions that produce CuIBr/HMTETA and yield polymers with narrow molecular weight distributions, which were previously not accessible from typical inifertermediated polymerizations. Alternatively, several other reports focused on controlling copper-catalyzed polymerizations without the need for an iniferter that would facilitate the polymerization. These systems employed various photoinitiators to directly adjust the relative amounts of CuII deactivator and CuI activator species in the reaction mixture upon exposure to light. In 2010, Yagci and coworkers demonstrated a UV switchable radical polymerization of MMA in methanol using CuIIBr2 ligated to N,N,N′,N″,N″pentamethyldiethylenetriamine (PMDETA) to mediate the reaction and ethyl 2-bromopropionate (EtBP) as an initiator.61 The authors utilized the ability of methanol to solvate the deactivating CuIIBr2/PMDETA complex and achieved a homogeneous polymerization. UV irradiation (λ = 350 nm) of the reaction mixture reduced the CuIIBr2/PMDETA deactivator complex to an active form (i.e., CuIBr/PMDETA), which induced radical formation via the homolysis of EtBP and initiated the growth of polymer chains. The presence of methanol also aided the oxidation of the CuI species as well as the reduction of

Scheme 17. Photoswitchable Combined Iniferter−ATRP-Type Polymerizations

1978

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

initiators, such as 2-bromopropionitrile (BPN) or EBiB, when exposed to visible light (λ = 350−550 nm).70 The polymerization method featured no photoinitiators; rather, visible light irradiation directly induced the reduction of the CuII deactivator to a CuI activator species which triggered polymer growth. After optimization of the type of initiator employed as well as the catalyst/ligand combination, a series of poly(MMA)s were obtained with relatively low molecular weight distributions (Đ = 1.12−1.56). The photoregulated character of the optimized system was later demonstrated to reversibly alter the rates of a CuIIBr2/TPMA-catalyzed polymerization of MMA via periodic exposure to visible light.70,71 In addition, the polymerization of MMA tolerated low levels of oxygen primarily because the visible light was highly effective in regenerating CuI species even after they were oxidized by O2.71 It was also shown that the generation of an active CuI catalyst was not limited to broad-band irradiation (λ = 350−550 nm) but was also possible upon exposure to specific wavelengths of light (i.e., λ = 366, 405, 408, 436, or 546 nm). Irradiation at single wavelengths (λ = 392 or 450 nm) and the use of a broad-band light source (i.e., sunlight) were also applied to selectively modulate Cu-mediated polymerizations. In 2012, Matyjaszewski and co-workers utilized light at the aforementioned wavelengths to successfully turn on the polymerization of various acrylate and methacrylate monomers.72 The authors further reported that irradiation at λ = 631 nm or the absence of light halted the reaction and yielded no polymeric products. The photoswitchable mechanism employed in this procedure was similar to that of the photoinitiator-free systems described above, where visible light was used to generate a CuI activator species from a deactivated CuII complex directly via the photoinduced homolytic cleavage of a Cu−Br bond. The process was significantly hindered in the absence of light but could be repeatedly restarted upon exposure to light of sufficient energy. The usefulness, durability, and simplicity of visible lightmodulated copper-mediated radical polymerization processes were also demonstrated through the use of inexpensive and widely accessible fluorescent lamps (λ = 400−750 nm). In a recent study, Jordan and co-workers utilized a fluorescent lamp light source in the CuIIBr2/PMDETA-catalyzed generation of brushes of poly(MMA).73 Upon exposure to light, the CuII species were successfully reduced. Similar to the systems mentioned previously, the CuI species generated in situ effectively facilitated the polymerization of MMA in the presence of an EBiB initiator. Turning off the fluorescent lamp hindered the polymerization of MMA, but subsequent re-exposure of the reaction mixture to photoirradiation restored the polymer chain growth. Ultimately, the process allowed for the formation of several homopolymers and block copolymer brushes which were grown from initiator-incorporated silicon dioxide surfaces consisting of self-assembled monolayers (SAMs) of (3aminopropyl)triethoxysilane (APTES) and 2-bromoisobutyryl bromide (BiBB).

Haddleton and co-workers further demonstrated that the use of photoinitiators was not required to achieve UV-regulated copper-mediated radical polymerizations.67 By utilizing a mixture of a CuIIBr2 catalyst with an aliphatic tertiary aminebased ligand (i.e., Me6TREN; TREN = tris(2-aminoethyl)amine) in the presence of various α-hydroxyl- and vic-diolfunctionalized initiators, temporal kinetic control over the radical polymerization of several different monomers was achieved. The authors proposed that the polymerization was started by the photoactivated free Me6TREN ligand, which reacted with the initiator via outer-sphere electron transfer to generate carbonbased radicals that initiated the chain propagation reaction. Styrene (St), MMA, and various acrylates were all successfully polymerized to high monomer conversion (>90%) after only 90 min of UV irradiation (λ ≈ 360 nm). Moreover, the polymerization process was effectively stopped in the absence of light, and subsequent re-exposure to UV irradiation restored the continued growth of polymer chains. Polymerizations were also induced upon exposure to sunlight, but required extended reaction times (15 h) to reach high conversion. Reflective of the excellent control bestowed by this switchable polymerization process, well-defined (Đ = 1.03−1.17) polymers were synthesized under optimized conditions. Later studies from the same group demonstrated almost quantitative monomer conversion with excellent control over polymer distribution (Đ ≈ 1.03) was also possible using only parts per million quantities of the preformed copper(II) formate catalyst [Cu(Me6TREN)(O2CH)](ClO4).68 This ionic complex in conjunction with EBiB was observed to promote the polymerization of various acrylates, such as methyl acrylate (MA), upon exposure to UV radiation (λ = 320−390 nm). In contrast, polymer chain growth ceased in the absence of light. The polymerization could be repeatedly hindered and restored via alternating exposures to UV radiation while preserving the high control over the resulting polymers that were produced (Đ ≈ 1.10). In comparison to the previous system, a photoactivated amine that initiated the polymerization upon UV irradiation was not required; instead, the formate component of the copper catalyst reduced the CuII species to an active CuI complex and directly facilitated the radical polymerization. 4.1.2. Visible Light Control. To avoid the need for specialized equipment necessary for UV-controlled coppermediated radical polymerizations, several groups have been able to utilize visible light to selectively turn off and on polymerization reactions. In 2014, Yagci and co-workers harvested the photoredox properties of dimanganese decacarbonyl (Mn2(CO)10) to induce and inhibit the copper-mediated radical polymerization of MMA, MA, and St upon alternating exposure to visible light.69 Under visible light irradiation (λ = 400−500 nm), Mn2(CO)10 underwent homolytic cleavage to generate • Mn(CO)5 radicals which later reduced the CuIIBr2/PMDETA complex. The resulting active CuI species, along with the alkyl halide initiator EtBP, promoted the controlled polymerization of MMA (Đ = 1.13−1.48). However, the growth of polymer chains was suspended in the absence of light since •Mn(CO)5 radicals did not appear to form under these conditions. Chain propagation was restored upon subsequent re-exposure to visible light, demonstrating the switchable character of the system. ́ In 2012, Mosnácě k and Ilčiková reported that low quantities (i.e., concentrations in the parts per million range) of a conventional CuIIBr2 catalyst ligated to PDMETA or TPMA could promote the polymerization of MMA in the presence of

4.2. Electrochemical Control

The wide range of possible triggers to enable tunable coppermediated radical polymerizations was demonstrated by Matyjaszewski and co-workers, who used an electrochemical stimulus to efficiently activate and deactivate the growth of polymer chains.74 Similar to the photoswitchable systems described above, the electrochemically switchable controlled polymerization was performed using a CuIIBr2/Me6TREN-type catalyst and initiated with EtBP in acetonitrile. The CuII species 1979

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

was successfully reduced to CuI under an applied potential of −0.69 V using platinum working and counter electrodes with a silver reference electrode, which ultimately initiated a polymerization reaction (Scheme 18). Adjusting the potential to −0.40 V

Scheme 19. General Mechanism of Photoredox-Mediated Radical Polymerizationsa

Scheme 18. Electrochemical Modulation of a Cu-Mediated Radical Polymerization

a

inhibited the polymer chain growth as the reduced potential was not sufficient to reduce the CuII species. Indeed, the shift from an anodic to a cathodic process halted the formation of CuI and effectively stopped the polymerization reaction. The electrochemically controlled radical polymerizations were amenable to in situ kinetic switching via the application of alternating potentials. Moreover, a series of well-defined (Đ = 1.06) polymers with tunable molecular weights were prepared via modulation of the applied potential. Zhou and co-workers later extended the electrochemically switchable technique to the synthesis of well-defined polymer brushes. Polymer chain growth on a gold surface was regulated by selectively altering the potential that was applied to a reaction mixture containing a CuIICl2/bpy (bpy = 2,2′-bipyridyl) catalyst system.75 The radical initiator-modified surface consisted of a gold-coated silicon wafer and a ω-mercaptoundecyl bromoisobutyrate-based initiator. With this system, 3-sulfopropyl methacrylate potassium salt (SPMA) monomers were polymerized from the surface to afford polymer brushes. The radical polymerization of SPMA proceeded in a controlled fashion for up to 5 h when a relatively low negative potential of −0.16 V was applied using a platinum gauze working electrode and a platinum wire counter electrode with a saturated calomel electrode (SCE) reference electrode. Moreover, polymer growth could be turned off by adjusting the applied potential to 0.02 V, which resulted in the accumulation of CuII species that ultimately terminated propagating radicals. When the potential was changed back to −0.16 V, the dormant CuII complexes were reduced to active CuI species which effectively restored the polymerization of SPMA.

PCat = photocatalyst and D = donor.

oxidation by the excited PCat*, or by hydrogen atom abstraction from the oxidized donor amines. In contrast, the oxidative pathway proceeds via direct reduction of an initiator species (PMn−I) by PCat* to generate carbon-based radicals that initiate polymerization reactions. The resulting PCat I− is a powerful oxidant and is capable of deactivating the propagating radical to regenerate the ground state photocatalyst (PCat). The polymerization can be reinitiated upon reduction of the now-dormant polymer chains by PCat*. While examples utilizing the reductive quenching pathway are known, the majority of reported photoredox-mediated polymerizations occur through the oxidative quenching mechanism. As the switchable behavior of this system is intrinsically limited to visible light irradiation, this section will be divided by catalyst type rather than stimulus type for clarity. 5.1. Transition Metal-Based Photoredox Catalysts

5.1.1. Ruthenium-Based Photoredox Catalysts. While the potential to utilize Ru-based photosensitizers as a means to reversibly initiate radical polymerizations was recognized as early as 1985,76 switchable polymerization activity was demonstrated only recently by Choi and co-workers.77 The combination of the RuII−polypyridine catalyst [Ru(bpy)3]Cl2 with a sacrificial amine donor (i.e., N,N-diisopropylethylamine) was found to generate carbon-based radicals via reduction of EBiB under visible light irradiation (λ > 420 nm). The resulting radical species initiated the free radical polymerization of various methacrylates, including MMA as well as ethyl (EMA), butyl (BMA), tertbutyl (TBMA), isobutyl (IBMA), glycidyl (GMA), and 2ethoxyethyl (EEMA) methacrylate, all of which proceeded to relatively high conversions. Catalyst loadings as low as 0.01 mol % were demonstrated to be effective, although the method was limited to methacrylate monomers and proved inefficient for St and acrylates. No monomer consumption was observed in the absence of the light source, allowing the polymerization of MMA to be effectively switched off and then back on (i.e., between relatively inactive and active states, respectively) multiple times over the course of a reaction by altering the photoirradiation process. While the polymerization was clearly light-driven and enabled switchable characteristics, the authors reported a gradual rise in dispersity with each switching cycle, likely due to irreversible chain termination occurring through hydrogen atom abstraction from the sacrificial donor. Despite the undesired termination, the number of residual active initiators was sufficient to induce the polymerization reaction and resulted in high monomer conversions even after multiple switching cycles. In parallel, Alexandrova and co-workers reported the polymerization of St, n-butyl acrylate (BA), and MMA using the

5. PHOTOREDOX-MEDIATED RADICAL POLYMERIZATION Photoredox catalysts are quickly becoming popular components in systems used to control reversible-deactivation radical polymerizations. Increased understanding of the photoredox process has led to a multitude of both organic- and transition metal-mediated photoredox polymerization techniques reported within the past few years. Though similar in scope and activity to Cu-mediated radical polymerizations, the divergent activation/ deactivation process merits a separate discussion. As shown in Scheme 19, visible light irradiation of a given photocatalyst (PCat) generates an excited state (PCat*) that can then participate in either reductive or oxidative quenching pathways. In reductive quenching, a sacrificial donor, which is typically a non-nucleophilic amine, reduces the excited catalyst to generate PCat•. This complex is a powerful reductant capable of homolyzing a carbon−initiator bond and initiating polymerization reactions. Propagating chains can be deactivated through 1980

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

ruthenium metallacycle cis-[Ru(ppy)(phen)(MeCN)2][PF6] (ppy = 2-pyridylphenyl; phen = phenanthroline).78 Subjecting the Ru precatalyst to visible light radiation (λ > 400 nm) resulted in the dissociation of one acetonitrile ligand and generated a reactive 16-electron intermediate. Oxidative halide addition of EBiB produced a carbon-centered radical species and initiated a polymerization reaction. The authors observed no polymerization activity in the absence of light and reported that solvent (i.e., acetone) coordination of the light-induced vacant site prevented further radical production. Additionally, reversible deactivation was accomplished through reductive halide transfer from the 17-electron Ru complex to the propagating radical species. Although low-dispersity (Đ = 1.22) poly(MMA) could be synthesized, the polymerizations of BA and St typically afforded polymers with relatively high molecular weight distributions (Đ = 1.5−1.6). The in situ switching of a homopolymerization of MMA by alternating exposure of the reaction mixture to photoirradiation was repeated multiple times, but only modest conversions were achieved after several switching cycles. The reduced activity was likely due to decomposition of the metastable solvento−Ru complex, which could irreversibly rearrange into thermodynamically more stable dimeric complexes, rendering the catalyst permanently inactive. In 2014, Boyer and co-workers extended the utility of [Ru(bpy)3]Cl2 as a photoredox initiator for radical polymerizations to a combined photoinduced electron transfer− reversible addition−fragmentation chain transfer (PET− RAFT) process.79 In combination with appropriate thiocarbonylthio RAFT initiators, [Ru(bpy)3]Cl2 was found to be an effective catalyst for the polymerization of MMA, MA, and N,Ndimethylacrylamide (DMA) under visible light irradiation (λ = 435 nm). All of these polymerizations resulted in high monomer conversions and yielded materials with narrow molecular weight distributions (Đ < 1.15). Moreover, despite a short induction period and slower propagation rates, the presence of oxygen had no effect on the outcome of the polymerization reactions. In this case, the photoexcited Ru complex produced via photoirradiation served a dual purpose: reduction of the thiocarbonylthio initiator to initiate the polymerization process along with coincident reduction of oxygen to nonparticipating superoxide. The polymerization of MMA could be selectively cycled between off (i.e., relatively inactive) and on (i.e., relatively active) states multiple times in the absence as well as in the presence of oxygen. This switchable system was further applied to a series of chain extension studies, where multiple additions of MA or DMA to a single reaction vessel resulted in the formation of successively larger polymers. Notably, a waiting period of one month in the absence of light before a second addition of monomer was demonstrated. Furthermore, the possibility to switch off the reaction at will was utilized to obtain multiblock copolymers of MA and DMA. These syntheses proceeded via sequential monomer additions and were facilitated by the removal of the light source at ca. 95% conversion for each individual block. The PET−RAFT technique was also extended to the polymerizations of various methacrylates, acrylates, and acrylamides in aqueous and biological media (i.e., in the presence of bovine serum albumin) to afford well-defined polymers with Đ values that were typically less than 1.3.80 The aqueous PET− RAFT of DMA was similarly switched between inactive and active states by oscillating between periods of darkness and photoirradiation (λ = 435 nm). Moreover, the system may offer the possibility of forming well-defined block copolymers from a variety of monomers outfitted with protein-based end groups.

Collectively, these results revealed that PET−RAFT facilitated by the visible light irradiation of [Ru(bpy)3]Cl2 is a versatile polymerization technique amenable to a variety of monomers and provides a high degree of control. More recently, Xu and Boyer reported a step-growth polymerization based on the thiol−ene reaction that was facilitated by [Ru(bpy)3]Cl2.81 As shown in Scheme 20, the Scheme 20. Photoredox Modulation of a Step-Growth Thiol−Ene Polycondensationa

a

ArNH2 = p-toluidine.

reaction proceeded through a reductive quenching pathway whereby p-toluidine transferred an electron to the photoexcited Ru complex (RuII*) to ultimately generate a thiyl radical species via hydrogen atom abstraction. The newly formed thiyl radical reacted with a comparatively electron-rich alkene, which in turn generated a carbon-based radical that was subsequently quenched by a free thiol and yielded a new thiyl radical. As opposed to the use of a traditional sacrificial donor, the anilinium species lost H+ to re-enter the catalytic cycle, while the ground state photocatalyst was regenerated by oxidative quenching from atmospheric oxygen. The combination of [Ru(bpy)3]Cl2 and ptoluidine with a dithiol and a diene resulted in a photoinitiated step-growth polycondensation when irradiated at λ = 461 nm. Though there were effectively three simultaneously operating redox cycles, the authors observed no polymerization activity in the absence of light and showed that the polymerization technique was switchable. Overall, the technique provides a highly controlled alternative to the traditional thermally-induced and UV-induced thiol−ene polycondensation methods. 5.1.2. Iridium-Based Photoredox Catalysts. The first switchable controlled/living radical polymerization mediated by an iridium-based photoredox catalyst was reported by Fors and Hawker in 2012.82 Visible light irradiation (using a fluorescent lamp) of the Ir catalyst fac-[Ir(ppy)3] in the presence of ethyl 2bromo-2-phenylacetate (EBPA) successfully generated a carbonbased radical through oxidative quenching, which consequently initiated the polymerization of MMA (Scheme 21). The polymerization proved to be highly tunable, with close agreement between theoretical and experimental molecular weights, and produced polymers with relatively low dispersity (Đ < 1.25). No monomer conversion was observed in the absence of light, and multiple switching cycles were demonstrated by repeatedly turning off the polymer chain growth upon the removal of the light source for 1 h intervals. Additionally, fac[Ir(ppy)3] also facilitated the polymerization of methacrylic acid (MAA), a difficult transformation for traditional ATRP methods. Hawker and co-workers later utilized this system in the controlled radical polymerization of various acrylates, including MA, BMA, and TBMA, and again demonstrated high monomer conversions to generate relatively well-defined polymers (Đ < 1981

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

Scheme 21. Generalized Mechanism of Photoredox-Mediated Radical Polymerizations Utilizing fac-Ir(ppy)3 as a Photocatalyst

Scheme 22. Combined PET−RAFT/Lewis Acid-Mediated Photoredox Polymerizations That Utilize fac-[Ir(ppy)3] and Y(OTf)3

1.5).83 In combination with benzyl α-bromoisobutyrate, fac[Ir(ppy)3] was shown to display switchable characteristics. Indeed, periodic removal of the light source allowed the polymerization reaction to be switched on and off (i.e., between relatively active and inactive states, respectively) multiple times, with no observable change in overall conversion or dispersity values. The incorporation of up to 50% acrylic acid (AA) in a random copolymerization with ethyl acrylate further demonstrated the robust nature of the fac-[Ir(ppy)3]-mediated polymerization method. In 2014, Boyer and co-workers reported fac-[Ir(ppy)3]mediated PET−RAFT (vide supra) of a multitude of monomers, including acrylates, methacrylates, acrylamides, and others.84 As in the [Ru(bpy)3]Cl2-mediated PET−RAFT process, oxidative quenching of the photoexcited (λ = 435 nm) catalyst by a thiocarbonylthio species initiated the polymerization reaction. All of the monomers investigated provided polymers with tunable molecular weights and low molecular weight distributions. Suspended chain propagation was observed in the absence of light, and a homopolymerization of MMA could be toggled between active and inactive states through intermittent exposure to visible light radiation. The switchable process was also utilized in the synthesis of a triblock copolymer by successive additions of MA, BA, and tert-butyl acrylate after periods of inactivity (i.e., in an off state). Additionally, the polymerizations of vinyl acetate (VAc) as well as N-vinylpyrrolidinone (NVP) could selectively be turned off and then back on (i.e., between relatively inactive and active states, respectively) by switching the light source accordingly, even in the presence of oxygen.85 The polymerization of NVP in the bulk was also investigated in the absence of oxygen and similar switching characteristics were observed. Later in 2014, the same group demonstrated the PET−RAFT of 3-oxobutyl acrylate (OBA) and 3-(trimethylsilyl)prop-2-yn-1yl acrylate (TMSPA) as well as acrylates containing alkyl chains of various lengths.86 A switchable fac-[Ir(ppy)3]-mediated PET− RAFT was again demonstrated by toggling the polymerization of n-hexyl acrylate between off (i.e., relatively inactive) and on (i.e., relatively active) states multiple times. In addition, the process was used to generate a novel triblock copolymer, poly(OBA-bMA-b-TMSPA). The polymerizations were allowed to proceed for 24 h after each successive monomer addition, while the light source was removed to deactivate the catalyst between additions. Most recently, Boyer and co-workers demonstrated stereocontrol in fac-[Ir(ppy)3]-mediated PET−RAFT polymerizations through the addition of a Lewis acid mediator.87 The combination of fac-[Ir(ppy)3] with Y(OTf)3 in the presence of a trithiocarbonate initiator facilitated the PET−RAFT of DMA to generate well-defined (Đ < 1.3), stereoregular polymers in high yield (Scheme 22). While atactic poly(DMA) was observed

in the absence of the Lewis acid, the addition of Y(OTf)3 ([Y(OTf)3]/[DMA] = 0.05) to the reaction mixture resulted in predominately isotactic (meso/rac = 0.83/0.17) polymers. Furthermore, the authors found that the addition of DMSO deactivated Y(OTf)3 and utilized this relationship to synthesize atactic-b-isotactic as well as isotactic-b-atactic stereoblock polymers. Remarkably, the controlled addition of DMSO to an ongoing polymerization resulted in the formation of stereogradient poly(DMA). The polymerization could be halted at a defined interval of forming isotactic poly(DMA) by removing the reaction mixture from light, at which point DMSO was added to deactivate a portion of the Y(OTf)3 catalyst. This process was repeated several times, which resulted in a stereogradient poly(DMA) that transitioned from isotactic toward an increasingly atactic stereochemistry. While the majority of effort directed toward Ir-based photoredox-mediated polymerizations has thus far focused on the combination of fac-[Ir(ppy)3] with α-haloester initiators, some recent studies have attempted to extend the utility of these systems through the use of novel initiators or other Ir-based photocatalysts. Yang and co-workers reported the use of perfluoro-1-iodohexane in conjugation with fac-[Ir(ppy)3] to initiate the polymerization of functional acrylates and access polymers with short perfluoroalkyl end groups.88 Additionally, Lalevée and co-workers synthesized the novel Ir complex Ir(btp)2(tmd) (btp = 2-(2′-benzothienyl)pyridine; tmd = 2,2,6,6-tetramethyl-3,5-heptanedione) and demonstrated its utility in the photoredox-mediated polymerizations of various acrylates.89 In both cases, the authors were able to show switchable activities similar to those described in the previous examples, where the polymerizations were effectively switched between off (i.e., relatively inactive) and on (i.e., relatively active) states by removing and replacing the light source, respectively. 5.1.3. Niobium-Based Photoredox Catalysts. Heterogeneous photoredox-mediated polymerization catalysts represent a relatively unexplored area. In 2014, Liu and co-workers reported the synthesis of surface-modified Nb(OH)5 nanoparticles and demonstrated their photoredox activity in RAFT polymerizations.90 The surface-grafting of benzyl alcohol to the Nb(OH)5 particles resulted in a graft density of roughly 0.114 mmol g−1 and a red shift of the absorbance of the material into the visible region. Visible light irradiation (λ > 420 nm) resulted in electron transfer to the Nb centers through ligand-to-metal charge transfer and generated oxygen-centered radical cations. Subsequent proton loss afforded relatively stable benzylic radicals, which then reacted with the trithiocarbonate RAFT agents present in solution and initiated the RAFT process. The 1982

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

reaction without significantly impacting monomer conversion or polymer dispersity. This metal-free method enabled the polymerization of various monomers with molecular weights and dispersities comparable to those of traditional Cu-catalyzed systems. Additionally, as opposed to traditional Ru- and Ir-based photoredox catalysts, the excited PTH was not adequately oxidizing to generate radicals from amines, allowing for the controlled polymerization of other monomers, such as 2(dimethylamino)ethyl methacrylate (DMAEMA). Theriot and co-workers also reported a metal-free ATRP system that utilized perylene as the organic photoredox catalyst.93 Exposure to visible light (using pure white lightemitting diodes (LEDs)) in the presence of EBPA was found to facilitate the polymerization of MMA. The molecular weights of the polymers produced were higher than predicted, and only modest yields could be achieved without significant increases in dispersity. However, despite these limitations, the authors were able to demonstrate the switchable properties of the system by oscillating the polymerization of MMA between off (i.e., relatively inactive) and on (i.e., relatively active) states in a fashion analogous to that of the PTH-mediated ATRP. More recently, Johnson and co-workers demonstrated a controlled radical polymerization method based on a trithiocarbonate iniferter by utilizing the PTH photoredox system.94 In the presence of visible light, the PTH/iniferter combination was used to initiate the polymerization of certain acrylamides and acrylates, and consistently produced polymers with excellent control over molecular weight and narrow dispersity values (Đ < 1.1). Removal of the light source deactivated propagating radicals to form a macroiniferter that could be reinitiated upon further visible light irradiation, allowing for the controlled reversible deactivation of the polymerization. In addition, the authors showed that the rate of the polymerization of NiPAAm could be regulated multiple times by periodic photoirradiation over the course of a single reaction without affecting the molecular weight distributions of the resulting polymers. In 2014, Boyer and co-workers enhanced their PET−RAFT polymerization method by investigating a variety of organic photoredox agents.95 Irradiation of several different organic dyes with blue LEDs (λ = 461 nm) in the presence of a thiocarbonylthio initiator and MMA revealed eosin Y to be the most efficient photoredox mediator. Proceeding via an oxidative quenching pathway, similar to the preceding example that involved a fac-[Ir(ppy)3]-mediated PET−RAFT, the polymerizations of MMA and several other monomers were shown to reach fairly high conversions while maintaining narrow dispersities (Đ < 1.3). As previously observed, radical generation was intrinsically linked to photoirradiation and allowed for temporal kinetic control over the polymerization of MMA. Eosin

method proved effective for the RAFT polymerization of several acrylamides and methacrylates, typically producing low-dispersity (Đ < 1.25) materials in decent yields. No monomer conversion was observed in the absence of light, and a single polymerization reaction of N-isopropylacrylamide (NiPAAm) was switched between inactive and active states by varying the light stimulus. The same group later reported an improved functionalization procedure and achieved a benzyl alcohol graft density of roughly 0.888 mmol g−1 by using NbCl5 nanoparticles as the starting material.91 Exposure of the functionalized nanoparticles to visible light radiation (λ > 420 nm) in the presence of EBiB efficiently initiated the ATRP-type polymerization of various acrylates in a manner similar to that described above. The system was again shown to display switchable activity in the polymerization of NiPAAm, although the conversion was somewhat diminished compared to that of the RAFT system. Simple centrifugation allowed for recovery of the Nb nanoparticles, and the authors demonstrated up to five recycles of the same catalyst without significant losses of control or activity. 5.2. Organic-Based Photoredox Catalysts

To expand beyond transition metal-based catalysts, Hawker and co-workers investigated the use of organic dyes as photoredox mediators for ATRP reactions.92 As shown in Scheme 23, visible Scheme 23. Photoredox-Mediated Organocatalyzed Radical Polymerization Utilizing PTH as a Photocatalyst

light irradiation (λ = 380 nm) of 10-phenylphenothiazine (PTH) produced a sufficiently reducing excited state to generate carbonbased radicals from a traditional ATRP initiator (i.e., EBPA). Initially, the polymerization of MMA was found to proceed in a controlled manner to furnish low-dispersity (Đ < 1.3) poly(MMA). Reversible deactivation was found to occur in a manner analogous to that of its transition metal-based analogues, allowing for the reaction to be reversibly switched to an inactive state by removing the light source. The authors demonstrated that the ATRP of benzyl methacrylate (BnMA) could be switched between off (i.e., relatively inactive) and on (i.e., relatively active) states up to six times over the course of a single

Scheme 24. Photoredox-Modulated PET−RAFT Polymerizations Promoted by Chlorophyll a or ZnTPP

1983

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

control over the polymerization reaction. It is therefore essential that radical production occurs in a reversible manner, allowing for an equilibrium between active and dormant species to be established and to persist throughout the entirety of the polymerization process. Photoinduced homolytic bond cleavage can satisfy these requirements in some cases and provides an alternative to the aforementioned methods used to initiate and control radical polymerizations.

Y-mediated PET−RAFT was also demonstrated to tolerate monomers containing functional groups that readily oxidize, such as DMAEMA. Remarkably, the authors also showed that the redox cycle could be switched to a reductive quenching process through the addition of trimethylamine, which acted as a sacrificial electron donor to render the system oxygen-tolerant. Building on these results, Boyer and co-workers subsequently employed chlorophyll a (Chl a) as a photoredox mediator of PET−RAFT.96 In a fashion analogous to that of photosynthesis, photon absorption generated an excited chlorophyll species (Chl a*) that could participate in an electron transfer reaction when an appropriate acceptor substrate was present (Scheme 24). The excited state produced by irradiation with blue or red light (λ = 461 or 635 nm) was capable of reducing a thiocarbonylthio initiator to initiate the PET−RAFT process. Similar to the group’s previous examples, a variety of acrylates, methacrylates, and acrylamides were polymerized with good molecular weight control and polymers with low dispersity values (Đ < 1.2) were obtained. The switchable behavior of the chlorophyll-mediated process was demonstrated via the repeated inhibition and reactivation of a homopolymerization of MA upon selective removal or introduction of the light source. Although Chl a is not strictly an organocatalyst, the porphyrin-coordinated Mg center is redox-neutral and did not appear to participate in the electron transfer process. Instead, the PET event originated from the excitation of the conjugated aromatic π-system of the porphyrin, and thus can be considered a nominally organocontrolled photoredox process. Boyer and co-workers later reported that the porphyrin-based switchable photoredox polymerization was also possible using the Zn−porhyrin complex (ZnTPP) shown in Scheme 24.97 Through a mechanism similar to that of the system described above, the metal center of ZnTPP did not partake in the electron transfer event during the activation of initiators. Instead, the authors proposed that Zn may coordinate the initiator compounds and facilitate activation. ZnTPP-mediated photoredox polymerization was shown to be most efficient with trithiocarbonate initiators and could be promoted with a range of different wavelengths (i.e., λ = 435−655 nm). Moreover, the rate of polymerization depended on the specific wavelength used: exposure to yellow light (λ = 565 nm) resulted in the fastest polymerizations followed by green and orange light (λ = 522 and 595 nm, respectively). The use of red or blue light (λ = 635 and 460 nm, respectively) afforded the lowest polymerization rates. Furthermore, only trace amounts of the ZnTPP catalyst (i.e., 50 ppm) were required to promote the polymerization of a range of monomers, including a variety of methacrylates and methacrylamides as well as St. Upon the exposure to red light, the polymerizations of various monomers yielded relatively welldefined (Đ = 1.04−1.40) polymers, even in the presence of oxygen. The polymerization of MA was also effectively turned between on (i.e., relatively active) and off (i.e., relatively inactive) states upon alteration between the presence and absence of blue or red light sources, demonstrating the switchable characteristic of this ZnTPP-mediated photoredox system.98

6.1. Photocontrol

6.1.1. Visible Light Control. In contrast to the aforementioned switchable copper-mediated polymerization developed by Yagci that utilized Mn2(CO)10 as a photoinitiator,69 an earlier report from Kamigaito and co-workers demonstrated that the Mn complex was a competent initiator for the controlled radical polymerizations of VAc, MA, and St.99 As previously discussed, visible light irradiation (λ > 400 nm) of Mn2(CO)10 resulted in homolytic cleavage of the Mn−Mn bond and generated Mn-based radicals. The newly formed radicals were shown to directly abstract iodine atoms from alkyl halide-based initiators to produce carbon-centered radicals, which initiated radical polymerizations. Moreover, the molecular weights of the resulting polymers obtained using this method were in relatively close agreement to the predicted values and displayed relatively low dispersity values (Đ ≥ 1.4). It was later shown that photoirradiation was essential to the polymerization process as no monomer conversions were observed in the absence of light. In addition, control reactions revealed that the Mn−I complex formed from the atom abstraction process was likely too stable to reversibly deactivate propagating radicals; as such, degenerative chain transfer primarily facilitated chain termination. This finding indicated that the switchable properties were due in part to iodine transfer from the Mn−I complex and were largely caused by slow Mn−Mn bond photolysis (i.e., a large proportion of the starting Mn2(CO)10 likely remained at the conclusion of the reaction). Regardless, the authors were able to demonstrate temporal control by switching the growth of polymer chains between off and on states multiple times over the course of a polymerization of VAc. Dimanganese decacarbonyl was also reported to be an efficient photoinitiator for the RAFT-type polymerization of MMA by Zhu, Zhu, and co-workers.100 The highly reactive Mn radical species generated by visible light irradiation (λ > 565 nm) of Mn2(CO)10 reduced a thiocarbonylthio RAFT agent which resulted in the generation of a carbon-based radical and initiated a RAFT process. Relatively high monomer conversion was achieved with good control over polymer molecular weight and dispersity (Đ < 1.2). The RAFT polymerization could also be initiated by exposing the reaction mixture to direct sunlight, although some minor broadening in the dispersity values was observed. In addition to MMA, MA, N-vinylcarbazole, and St were well tolerated by this method but typically resulted in lower monomer conversions. Similar to the system described by Kamigaito, no monomer conversion was observed in the absence of light. In addition, the polymerization rates of MMA were successfully toggled between on (i.e., relatively active) and off (i.e., relatively inactive) states through periodic exposure to light. In 2009, Nakamura and co-workers reported an organotellurium-mediated controlled/living radical polymerization of various acrylates, acrylamides, and other monomers.101 Visible light irradiation (λ > 470 nm) of ethyl 2-(phenyltellanyl)-2methylpropionate resulted in homolytic cleavage of the C−Te bond and directly generated tertiary carbon-centered radicals

6. RADICAL POLYMERIZATION VIA CONTROLLED BOND PHOTOLYSIS As previously mentioned, a fundamental aspect of reversibledeactivation radical polymerizations is the ability to control the concentration of propagating radical species. A rapid onedirectional generation of radicals would lead to chaotic initiation, propagation, and termination processes, and result in a loss of 1984

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

Scheme 25. Switchable Polymerizations Facilitated by Reversible Co−C Bond Cleavage in a Co−Porphyrin Complex

by cycling periods of irradiation multiple times over the course of a single IEAM polymerization. Moreover, by removing the light source after approximately 50% monomer conversion, the authors took advantage of the suspended propagation and formed novel homo-random block copolymers by the addition of a second monomer prior to subsequent photoirradiation. In 2012, Kaji, Goto, and co-workers developed a new variant of photolysis-induced controlled radical polymerization, coined reversible complexation-mediated polymerization (RCMP).106 This system utilized only an alkyl iodide initiator and an aminebased catalyst to initiate the polymerization of various acrylate monomers. The authors proposed the contribution of two different mechanisms. Either the irradiation caused homolytic bond cleavage of a C−I bond and the released I• was captured by the amine-based catalyst, or an amine−initiator complex was initially formed followed by C−I bond dissociation. The resulting unstable amine−iodine radical complex likely recombined with a second complex to give an amine-complexed iodine (I2) species. Despite the uncertainty of the initiation mechanism, subjecting a reaction mixture containing an alkyl iodide and tributylamine to visible light radiation (λ = 350−600 nm) promoted the polymerization of a variety of acrylates to yield relatively welldefined (Đ < 1.4) polymers. No monomer conversion was observed in the absence of light, indicating the formed amine complex efficiently quenched the propagating radicals. Periodic suppression and restoration of the polymer chain growth of MMA was achieved by the removal and reintroduction of light, and thus demonstrated the reversible nature of the deactivation process. Recently, the same group improved their RCMP methodology to utilize light with a wide range of wavelengths (λ = 350−750 nm) by employing an array of organic dyes as photocatalysts.107 Two different cyanine dyes, CY1 and CY2, as well as the spiropyran dye SP1 were each shown to have distinct absorption maxima, even in the presence of an alkyl iodide initiator, which could be used to control catalytic activity (Scheme 26a). While mechanistically similar to the tributylamine-mediated RCMP (vide supra), the authors reported that the photocatalysts acted as antennae that transferred light energy to the initiator species and caused C−I bond homolysis. Each of the three photocatalysts produced well-defined (Đ < 1.3) poly(MMA) in good yields. Moreover, all three catalysts could be activated independently of each other at different wavelengths, representing an additional handle for controlling polymerization activity. Consequently, instead of the removal of the light source necessary in the previous examples, the polymerization of MMA could be reversibly switched between off (i.e., relatively inactive) and on (i.e., relatively active) states by toggling the irradiation between higher and lower wavelengths, depending on the catalyst employed. The ability to halt propagation while still irradiating the solution imparts additional functionality to the system, whereby the combination of two different photocatalysts, each absorbing at different wavelengths, could facilitate two different polymerization reactions. Indeed, by employing the bifunctional initiator

that initiated radical polymerizations. All of the monomers explored were polymerized in high conversions, and well-defined polymers with narrow dispersities (Đ < 1.2), with the exception of poly(AA) (Đ = 1.34), were obtained. In addition, no polymerization activity was observed in the absence of light, indicating the Te-centered radical produced upon photolysis was capable of deactivating the propagating radical chains. This reversible deactivation process was utilized to repeatedly oscillate a BA polymerization between active and inactive states by turning the light source on and off, with no observed loss in control or dispersity of the resulting polymers produced. Similarly, in 2013, Fu and co-workers examined cobalt− porphyrin macroinitiators for the controlled radical polymerization of acrylamides.102 As shown in Scheme 25, poly(MA)- or poly(DMA)-appended cobalt−porphyrin complexes were found to undergo C−Co homolytic bond scission upon exposure to visible light radiation (λ = 400−800 nm). The liberated carbonbased radical directly initiated the polymerization of DMA, as well as other functionalized acrylamides, to furnish low-dispersity (Đ < 1.3) polymers in high yield. Reversible chain deactivation was also shown to occur via radical recombination with the Cocentered radical, which allowed for temporal control over the polymerization rates of DMA through alternate exposure to a light stimulus. In a subsequent report, the authors found that acrylate monomers (i.e., MA, BA, TBA) were also well tolerated by this system.103 Again, no polymerization activity was observed in the absence of light, and multiple in situ switches of the polymerization kinetics of BA were demonstrated. Extending beyond transition metal species, Cai and coworkers developed an aqueous RAFT-type polymerization of acrylic monomers promoted by the C−P bond photolysis of (2,4,6-trimethylbenzoyl)diphenylphosphine oxide (TPO).104 Visible light irradiation (λ = 420 nm) of TPO, in the presence of a thiocarbonylthio RAFT initiator, mediated the polymerization of functionalized acrylates, such as N-[2-(acryloyloxy)ethyl]pyrrolidinone (NAP) and poly(ethylene glycol) methyl ether acrylate (PEGA). The photolysis-produced P-centered radicals were shown to deactivate propagating radicals and resulted in highly controlled polymerization reactions that afforded polymers with low dispersities (Đ < 1.2); subsequent exposure to visible light reinitiated the deactivated chains. Moreover, alternating exposure of the reaction mixture to the light source enabled the NAP polymerization reaction to be toggled between different states of activity. The same group demonstrated that using sodium phenyl(2,4,6-trimethylbenzoyl)phosphinate (SPTP) as the initiator enabled photoswitchable control over the RAFT-type polymerization of ionic acrylamides in acidic media (pH 2.8).105 Visible light (λ = 420 nm) photolysis of SPTP in the presence of a thiocarbonylthio RAFT initiator induced the well-controlled RAFT polymerization of N-(2-aminoethyl)acrylamide hydrochloride (AEAM) as well as [2-(4-imidazolyl)ethyl]acrylamide (IEAM). As previously observed, propagation was suspended in the absence of light, indicating reversible deactivation of the radical chain ends. Temporal control was further demonstrated 1985

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

different polymerization mechanisms may lead to new classes of copolymers with advanced microstructures. 6.1.2. Ultraviolet Light Control. While visible lightinduced bond dissociation has received significant attention, some reports describe the exploration of UV-regulated bond photolysis. In 2013, Johnson and co-workers developed a bisnorbornene-functionalized iniferter that promoted the controlled radical polymerization of NiPAAm.108 Exposure of the trithiocarbonate iniferter to UV light (λ = 350 nm) resulted in homolytic C−S bond scission and generated a carbon-centered radical that initiated the polymerization of NiPAAm. Reversible chain deactivation via radical recombination led to poly(NiPAAm) with a narrow molecular weight distribution (Đ = 1.1) and high end group fidelity. No monomer consumption was observed in the absence of light, allowing the polymerization to be switched between inactive and active states by turning the light off and on, respectively. Direct exposure to sunlight also initiated the polymerization of NiPAAm without observable losses in monomer conversion or control. Additionally, the telechelic structure of the poly(NiPAAm) produced from this method was demonstrated by the selective modification of the norbornene end groups through a Diels−Alder reaction with a tris-tetrazine diene and ultimately afforded access to three-arm star polymers (Scheme 27). Wolpers and Vana also reported the UV-mediated controlled radical polymerization of several methacrylates.109 UV irradiation (λ = 366 nm) of an alkyl iodide chain transfer agent in the presence of AIBN generated carbon-based radicals through the homolysis of C−I bonds and initiated the polymerization of BMA. Deactivation of the radical chain ends was observed in the absence of light, whereas the polymer propagation could be reinitiated upon further irradiation. A single polymerization reaction of BMA could be switched between these off (i.e., relatively inactive) and on (i.e., relatively active) states multiple times and still produced relatively well-defined (Đ < 1.5) polymers in good yields.

Scheme 26. (a) Mechanism of Reversible ComplexationMediated Polymerization with Various Organic Photocatalysts and (b) Synthesis of Block Copolymers via Selective Activation and Deactivation of Multiple Polymerization Processes

35 in combination with catalyst SP1, a triarylsulfonium photoacid generator (PAG), MMA, and VL, the authors were able to demonstrate a one-pot synthesis of a block copolymer by independently controlling two different polymerization reactions (Scheme 26b). First, subjecting the reaction mixture to visible light radiation (λ = 550−750 nm) resulted in 67% conversion of MMA to polymer, while no conversion of VL could be observed, consistent with selective activation of the RCMP process. Subsequent tuning of the irradiation wavelength (λ = 350−380 nm) initiated the ROP of VL and resulted in 80% monomer conversion with only 1% further conversion of MMA, and resulted in the formation of a well-defined (Đ = 1.2) diblock copolymer. The ability to use orthogonal irradiation techniques to selectively control the synthesis of block copolymers from

Scheme 27. Radical Polymerization of NiPAAm by Photoinduced Bond Homolysis of an Iniferter Followed by End Group Postfunctionalization To Form Star Polymers

1986

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

Scheme 28. Random Copolymer Formation via Selective Monomer Discrimination in a Bichromophoric Transition MetalCatalyzed Polymerization

7. ALTERNATIVE POLYMERIZATION METHODS 7.1. Photocontrolled Olefin Polymerization

Palladium-based catalysts have also been employed to facilitate the polymerization of terminal olefins, but often proceed through nonradical pathways. The Pd-catalyzed polymerizations typically begin with the coordination of the monomer to the Pd active site, followed by insertion of the coordinated monomer into a Pd− alkyl bond, which induces polymer chain growth. Until recently, switchable nonradical olefin polymerization has been a largely unexplored area of controlled polymerization methods. However, in 2015, Akita, Inagaki, and co-workers developed visible light-controlled homo- and copolymerizations of terminal olefins based on a bichromophoric cationic iridium−palladium complex (36) (Scheme 28).110 The authors demonstrated that, in the absence of light, the PdII center successfully catalyzed the polymerization of 2,2,2-trifluoroethyl vinyl ether (TFEVE) but was inactive toward the polymerization of St, yielding only the dimeric species 37 instead. Upon exposure to visible light radiation (λ > 420 nm), the PdII species facilitated the polymerization of both TFEVE and St to afford relatively welldefined (Đ = 1.4−1.5) random poly(TFEVE/St) copolymers. It should be noted that the polymerization of TFEVE was not halted during the irradiation period; rather, it was proposed that photoexcitation of the PdII center by the IrIII chromophore led to diminished activation energy for the Pd−alkyl insertion step of the corresponding catalytic cycle. As a result, the polymerization of the previously inactive monomer (i.e., St) was enabled in addition to the polymerization of the more favored TFEVE monomer. Although the polymerization of TFEVE proceeded with no incorporation of St in the absence of light, visible light irradiation resulted in almost equivalent (1:1.15) incorporation of TFEVE and St after 48 h. The switchable Pd-catalyzed olefin polymerization method offered excellent tunability of the polymerization process and produced relatively well-defined (Đ = 1.5) copolymers.

Figure 2. (A) FK- and FE-based dendritic peptides and (B) their pHcontrolled supramolecular assembly into homo- and copolymers. Adapted with permission from ref 111. Copyright 2015 Wiley-VCH.

to the attractive Coulomb interactions formed between the FK and FE monomeric units. A subsequent switch to acidic conditions (pH < 4) resulted in protonation of the glutamic acid components of FE, effectively eliminating the electrostatic interactions between the FK and FE species, and afforded selfassembled FE homopolymers. In contrast, increasing the solution pH (pH > 10) deconstructed the self-assembled FE homopolymers and deprotonated the lysine components of the FK peptides. The hydrophobic interactions formed between the deprotonated dendritic peptides ultimately forced the selfassembly of the FK homopolymers. The use of a pH switch to enable reversible structural alterations of supramolecular polymer compositions represents a novel approach toward switchable polymerization systems and may lead to new classes of self-assembled systems.

7.2. pH-Controlled Supramolecular Polymerization

In a recent report, Besenius and co-workers described a pHregulated system for the selective and reversible formation of supramolecular homo- and copolymers of dendritic peptides.111 As shown in Figure 2A, phenylalanine−lysine (FK)-based and phenylalanine−glutamic acid (FE)-based amphiphilic peptides coupled to hydrophilic tetraethylene glycol peripheral dendron groups were utilized as macromonomers. Exposure of the mixture of FK and FE dendritic peptides to nearly neutral buffered conditions (pH 7.4) resulted in the formation of supramolecular FK−FE copolymers (Figure 2B). The spontaneous arrangement into supramolecular structures was attributed

8. SUMMARY AND OUTLOOK Various forms of external stimuli have been shown to be effective for the in situ regulation of a broad range of polymerization techniques. Photoirradiation and redox chemistry have been the 1987

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

most widely investigated “switches” thus far; however, other stimuli, including acid−base interactions, mechanochemical force, electrochemical potential, and changes in temperature, have been successfully employed to reversibly modulate the rate, stereochemical preference, and/or other outcomes of various polymerization reactions. Although many of the reported switchable polymerization catalysts have only been explored for relatively straightforward on/off kinetic control, some recent examples have begun to venture into more complex switching behaviors, such as the selective polymerization of specific monomers within various mixtures to control block copolymer architectures. In the past decade, a multitude of reports have emerged providing valuable fundamental insight into switchable polymerization methodologies, but the limited number of available techniques with advanced switchable activity highlights the infancy of the field as a whole. While significant progress has clearly been made, future efforts should be directed toward achieving more sophisticated control over the switchable characteristics of a system. For instance, the synthesis of complex materials with precisely tailored microstructures may be possible through the development of a polymerization catalyst capable of two discrete switching mechanisms induced by orthogonal external stimuli. Alternatively, the development of novel switchable polymerization catalysts containing multiple distinct catalytic centers that can be selectively switched on and off independent of each other could also lead to an enhanced level of control over polymer architecture (e.g., tacticity, composition, topology, etc.). Nevertheless, the nascent field of switchable polymerizations represents a useful alternative to conventional methods and is expected to facilitate the realization of new polymeric materials that are currently challenging or impossible to synthesize using known techniques.

chemical reactivity, specifically through the design of photoswitchable NHC scaffolds.

Dominika N. Lastovickova received B.A. degrees in chemistry and psychology from the College of the Holy Cross in 2012, after which she joined the laboratory of Prof. Bielawski at the University of Texas at Austin. Her research interests focus on the development of redoxtunable polymerization catalysts as well as the investigation of carbenefacilitated transformations.

AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Author Contributions ∥

Christopher W. Bielawski received a B.S. degree in chemistry from the University of Illinois, Urbana-Champaign, and a Ph.D. degree from the California Institute of Technology. After a postdoctoral appointment (also at Caltech), he assumed an independent position at the University of Texas at Austin where he directed synthetic efforts in a broad range of polymer and materials chemistry projects. Recently, he moved his research program to UNIST and is currently participating in a new initiative focused on the synthesis and study of novel macromolecular materials.

A.J.T. and D.N.L. contributed equally to this work.

Notes

The authors declare no competing financial interest. Biographies

ACKNOWLEDGMENTS We are grateful to the Office of Naval Research (Grant N0001414-1-0650), the Institute for Basic Science (Grant IBS-R019D1), and the BK21 Plus Program as funded by the Ministry of Education and the National Research Foundation of Korea for their support. ABBREVIATIONS AA acrylic acid AEAM N-(2-aminoethyl)acrylamide hydrochloride AgOTf silver trifluoromethanesulfonate AIBN azobis(isobutyronitrile) APTES (3-aminopropyl)triethoxysilane ATRP atom transfer radical polymerization

Aaron J. Teator received a B.S. degree in chemistry from the University of Nevada, Reno, in 2012. After his undergraduate studies, Aaron joined the research group of Prof. Bielawski at the University of Texas at Austin. His research interests focus on the development of externally switchable 1988

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

n-butyl acrylate tetrakis[3,5-bis(trifluoromethyl)phenyl]borate 2-bromoisobutyryl bromide butyl methacrylate benzyl methacrylate 2-bromopropionitrile 2,2′-bipyridyl 2-(2′-benzothienyl)pyridine cyclodextrin chlorophyll a cinnamoyl ε-caprolactone cobaltocene cis,cis-1,5-cyclooctadiene controlled radical polymerization copper(I)-catalyzed azide alkyne cycloaddition polydispersity index 1,8-diazabicyclo[5.4.0]undec-7-ene 2,3-dichloro-5,6-dicyano-1,4-benzoquinone density functional theory N,N-dimethylacrylamide 2-(dimethylamino)ethyl methacrylate ethyl α-bromoisobutyrate ethyl 2-bromo-2-phenylacetate 2-ethoxyethyl methacrylate ethyl methacrylate 2-(N,N-diethyldithiocarbamyl)isobutyric acid ethyl ester EtBP ethyl 2-bromopropionate FcPF6 ferrocenium hexafluorophosphate FeCP2 ferrocene GMA glycidyl methacrylate HMTETA hexamethyltriethylenetetramine IBMA isobutyl methacrylate IEAM [2-(4-imidazolyl)ethyl]acrylamide Irgacure 1-[4-(2-hydroxyethoxy)phenyl]-2-hydroxy-2methyl-1-propan-1-one L-LA L-lactide MA methyl acrylate MAA methacrylic acid meso-LA meso-lactide MIM N-methylimidazole MMA methyl methacrylate NaHMDS sodium hexamethyldisilazide NAP N-[2-(acryloyloxy)ethyl]pyrrolidinone NHC N-heterocyclic carbene NIPAAm N-isopropylacrylamide NVP N-vinylpyrrolidinone OBA 3-oxobutyl acrylate OEOMA oligo(ethylene glycol) monomethyl ether methacrylate PCat photocatalyst PEGA poly(ethylene glycol) methyl ether acrylate PET−RAFT photoinduced electron transfer−reversible addition−fragmentation chain transfer phen phenanthroline PMDETA N,N,N′,N″,N″-pentamethyldiethylenetriamine ppy 2-phenylpyridine PTH 10-phenylphenothiazine rac-LA rac-lactide RAFT reversible addition−fragmentation chain transfer RCMP reversible complexation-mediated polymerization ROMP ring-opening metathesis polymerization

ROP SAMs SCE SPMA SPTP

BA BArF BiBB BMA BnMA BPN bpy btp CD Chl a CIO CL CoCP2 COD CRP CuAAC Đ DBU DDQ DFT DMA DMAEMA EBiB EBPA EEMA EMA EMADC

St TBD TBMA TFEVE tmd TMSPA TPMA TPO tpp TREN VAc VL

ring-opening polymerization self-assembled monolayers saturated calomel electrode 3-sulfopropyl methacrylate potassium salt sodium phenyl(2,4,6-trimethylbenzoyl)phosphinate styrene 1,5,7-triazabicyclododecene tert-butyl methacrylate 2,2,2-trifluoroethyl vinyl ether 2,2,6,6-tetramethyl-3,5-heptanedione 3-(trimethylsilyl)prop-2-yn-1-yl acrylate tris(2-pyridylmethyl)amine (2,4,6-trimethylbenzoyl)diphenylphosphine oxide tetraphenylporphyrin tris(2-aminoethyl)amine vinyl acetate δ-valerolactone

REFERENCES (1) Malmström, E. E.; Hawker, C. J. Macromolecular engineering via ‘living’ free radical polymerizations. Macromol. Chem. Phys. 1998, 199, 923−935. (2) Matyjaszewski, K. Macromolecular engineering by controlled/ living ionic and radical polymerizations. Macromol. Symp. 2001, 174, 51−68. (3) Matyjaszewski, K. Macromolecular engineering: From rational design through precise macromolecular synthesis and processing to targeted macroscopic material properties. Prog. Polym. Sci. 2005, 30, 858−875. (4) Hawker, C. J.; Wooley, K. L. The Convergence of Synthetic Organic and Polymer Chemistries. Science 2005, 309, 1200−1205. (5) Matyjaszewski, K.; Tsarevsky, N. V. Macromolecular Engineering by Atom Transfer Radical Polymerization. J. Am. Chem. Soc. 2014, 136, 6513−6533. (6) Webster, O. W. Living Polymerization Methods. Science 1991, 251, 887−893. (7) Yagci, Y.; Tasdelen, M. A. Mechanistic transformations involving living and controlled/living polymerization methods. Prog. Polym. Sci. 2006, 31, 1133−1170. (8) Lutz, P. J. Control of macromolecular architectures by various polymerization methods: advantages and drawbacks. Macromol. Symp. 2000, 161, 53−62. (9) Hadjichristidis, N.; Iatrou, H.; Pitsikalis, M.; Mays, J. Macromolecular architectures by living and controlled/living polymerizations. Prog. Polym. Sci. 2006, 31, 1068−1132. (10) Stoll, R. S.; Hecht, S. Artificial Light-Gated Catalyst Systems. Angew. Chem., Int. Ed. 2010, 49, 5054−5075. (11) Lüning, U. Switchable Catalysis. Angew. Chem., Int. Ed. 2012, 51, 8163−8165. (12) Neilson, B. M.; Bielawski, C. W. Illuminating Photoswitchable Catalysis. ACS Catal. 2013, 3, 1874−1885. (13) Blanco, V.; Leigh, D. A.; Marcos, V. Artificial Switchable Catalysts. Chem. Soc. Rev. 2015, 44, 5341−5370. (14) Yagci, Y.; Reetz, I. Externally stimulated initiator systems for cationic polymerization. Prog. Polym. Sci. 1998, 23, 1485−1538. (15) Naumann, S.; Buchmeiser, M. R. Latent and Delayed Action Polymerization Systems. Macromol. Rapid Commun. 2014, 35, 682−701. (16) Naumann, S.; Buchmeiser, M. R. Liberation of N-heterocyclic carbenes (NHCs) from thermally labile progenitors: protected NHCs as versatile tools in organo- and polymerization catalysis. Catal. Sci. Technol. 2014, 4, 2466−2479. (17) Tzur, E.; Lemcoff, N. G. Latent Ruthenium Olefin Metathesis Catalysts for Ring-Opening Metathesis Polymerization (ROMP). In Handbook of Metathesis (set); Grubbs, R. H., Wenzel, A. G., O’Leary, D. 1989

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

J., Khosravi, E., Eds.; Wiley-VCH: Weinheim, Germany, 2015; pp 283− 312. (18) Yoon, T. P.; Ischay, M. A.; Du, J. Visible light photocatalysis as a greener approach to photochemical synthesis. Nat. Chem. 2010, 2, 527− 532. (19) Narayanam, J. M. R.; Stephenson, C. R. J. Visible light photoredox catalysis: applications in organic synthesis. Chem. Soc. Rev. 2011, 40, 102−113. (20) Xuan, J.; Xiao, W.-J. Visible-Light Photoredox Catalysis. Angew. Chem., Int. Ed. 2012, 51, 6828−6838. (21) May, P. A.; Moore, J. S. Polymer mechanochemistry: techniques to generate molecular force via elongational flows. Chem. Soc. Rev. 2013, 42, 7497−7506. (22) Brantley, J. N.; Wiggins, K. M.; Bielawski, C. W. Polymer mechanochemistry: the design and study of mechanophores. Polym. Int. 2013, 62, 2−12. (23) Yagci, Y.; Jockusch, S.; Turro, N. J. Photoinitiated Polymerization: Advances, Challenges, and Opportunities. Macromolecules 2010, 43, 6245−6260. (24) Vidavsky, Y.; Lemcoff, N. G. Light-induced olefin metathesis. Beilstein J. Org. Chem. 2010, 6, 1106−1119. (25) Yamago, S.; Nakamura, Y. Recent progress in the use of photoirradiation in living radical polymerization. Polymer 2013, 54, 981−994. (26) Leibfarth, F. A.; Mattson, K. A.; Fors, B. P.; Collins, H. A.; Hawker, C. J. External Regulation of Controlled Polymerizations. Angew. Chem., Int. Ed. 2013, 52, 199−210. (27) Chatani, S.; Kloxin, C. J.; Bowman, C. N. The power of light in polymer science: photochemical processes to manipulate polymer formation, structure, and properties. Polym. Chem. 2014, 5, 2187−2201. (28) Moad, G.; Keddie, D.; Guerrero-Sanchez, C.; Rizzardo, E.; Thang, S. H. Advances in Switchable RAFT Polymerization. Macromol. Symp. 2015, 350, 34−42. (29) Breslow, D. S. Metathesis Polymerization. Prog. Polym. Sci. 1993, 18, 1141−1195. (30) Bielawski, C. W.; Grubbs, R. H. Living ring-opening metathesis polymerization. Prog. Polym. Sci. 2007, 32, 1−29. (31) P'Pool, S. J.; Schanz, H.-J. Reversible Inhibition/Activation of Olefin Metathesis: A Kinetic Investigation of ROMP and RCM Reactions with Grubbs’ Catalysts. J. Am. Chem. Soc. 2007, 129, 14200−14212. (32) Peeck, L. H.; Leuthäusser, S.; Plenio, H. Switched Stereocontrol in Grubbs-Hoveyda Complex Catalyzed ROMP Utilizing ProtonSwitched NHC Ligands. Organometallics 2010, 29, 4339−4345. (33) Balof, S. L.; Yu, B.; Lowe, A. B.; Ling, Y.; Zhang, Y.; Schanz, H.-J. Ru-Based Olefin Metathesis Catalysts Bearing pH-Responsive NHeterocyclic Carbene (NHC) Ligands: Activity Control via Degree of Protonation. Eur. J. Inorg. Chem. 2009, 2009, 1717−1722. (34) Piermattei, A.; Karthikeyan, S.; Sijbesma, R. P. Activating catalysts with mechanical force. Nat. Chem. 2009, 1, 133−137. (35) For a photocontrolled ROMP system that operates via a similar mechanism, see: Diesendruck, C. E.; Vidavsky, Y.; Ben-Asuly, A.; Lemcoff, N. G. A Latent S-Chelated Ruthenium Benzylidene Initiator for Ring-Opening Metathesis Polymerization. J. Polym. Sci., Part A: Polym. Chem. 2009, 47, 4209−4213. (36) Savka, R.; Foro, S.; Gallei, M.; Rehahn, M.; Plenio, H. OxidationTriggered Ring-Opening Metathesis Polymerization. Chem. - Eur. J. 2013, 19, 10655−10662. (37) Varnado, C. D., Jr.; Rosen, E. L.; Collins, M. S.; Lynch, V. M.; Bielawski, C. W. Synthesis and study of olefin metathesis catalysts supported by redox-switchable diaminocarbene[3]ferrocenophanes. Dalton Trans. 2013, 42, 13251−13264. (38) Rosen, E. L.; Varnado, C. D., Jr.; Arumugam, K.; Bielawski, C. W. Metal-centered oxidations facilitate the removal of ruthenium-based olefin metathesis catalysts. J. Organomet. Chem. 2013, 745−746, 201− 205. (39) For other examples of redox-switchable olefin metathesis catalysts that may find utility in controlling ROMP reactions, see: Arumugam, K.; Varnado, C. D., Jr.; Sproules, S.; Lynch, V. M.; Bielawski, C. W. Redox-

Switchable Ring-Closing Metathesis: Catalyst Design, Synthesis, and Study. Chem. - Eur. J. 2013, 19, 10866−10875. (40) Ogawa, K. A.; Goetz, A. E.; Boydston, A. J. Metal-Free RingOpening Metathesis Polymerization. J. Am. Chem. Soc. 2015, 137, 1400−1403. (41) Dechy-Cabaret, O.; Martin-Vaca, B.; Bourissou, D. Controlled Ring-Opening Polymerization of Lactide and Glycolide. Chem. Rev. 2004, 104, 6147−6176. (42) Kamber, N. E.; Jeong, W.; Waymouth, R. M.; Pratt, R. C.; Lohmeijer, B. G. G.; Hedrick, J. L. Organocatalytic Ring-Opening Polymerization. Chem. Rev. 2007, 107, 5813−5840. (43) Thomas, C. M. Stereocontrolled ring-opening polymerization of cyclic esters: synthesis of new polyester microstructures. Chem. Soc. Rev. 2010, 39, 165−173. (44) Guillaume, S. M.; Kirillov, E.; Sarazin, Y.; Carpentier, J.-F. Beyond Stereoselectivity, Switchable Catalysis: Some of the Last Frontier Challenges in Ring-Opening Polymerization of Cyclic Esters. Chem. Eur. J. 2015, 21, 7988−8003. (45) Coulembier, O.; Dove, A. P.; Pratt, R. C.; Sentman, A. C.; Culkin, D. A.; Mespouille, L.; Dubois, P.; Waymouth, R. M.; Hedrick, J. L. Latent, Thermally Activated Organic Catalysts for the On-Demand Living Polymerization of Lactide. Angew. Chem., Int. Ed. 2005, 44, 4964−4968. (46) Coulembier, O.; Lohmeijer, B. G. G.; Dove, A. P.; Pratt, R. C.; Mespouille, L.; Culkin, D. A.; Benight, S. J.; Dubois, P.; Waymouth, R. M.; Hedrick, J. L. Alcohol Adducts of N-Heterocyclic Carbenes: Latent Catalysts for the Thermally-Controlled Living Polymerization of Cyclic Esters. Macromolecules 2006, 39, 5617−5628. (47) Osaki, M.; Takashima, Y.; Yamaguchi, H.; Harada, A. Switching of polymerization activity of cinnamoyl-α-cyclodextrin. Org. Biomol. Chem. 2009, 7, 1646−1651. (48) Neilson, B. M.; Bielawski, C. W. Photoswitchable NHCpromoted ring-opening polymerizations. Chem. Commun. 2013, 49, 5453−5455. (49) Gregson, C. K. A.; Gibson, V. C.; Long, N. J.; Marshall, E. L.; Oxford, P. J.; White, A. J. P. Redox Control within Single-Site Polymerization Catalysts. J. Am. Chem. Soc. 2006, 128, 7410−7411. (50) Broderick, E. M.; Guo, N.; Vogel, C. S.; Xu, C.; Sutter, J.; Miller, J. T.; Meyer, K.; Mehrkhodavandi, P.; Diaconescu, P. L. Redox Control of a Ring-Opening Polymerization Catalyst. J. Am. Chem. Soc. 2011, 133, 9278−9281. (51) Wang, X.; Thevenon, A.; Brosmer, J. L.; Yu, I.; Khan, S. I.; Mehrkhodavandi, P.; Diaconescu, P. L. Redox Control of Group 4 Metal Ring-Opening Polymerization Activity toward L-Lactide and εCaprolactone. J. Am. Chem. Soc. 2014, 136, 11264−11267. (52) Broderick, E. M.; Guo, N.; Wu, T.; Vogel, C. S.; Xu, C.; Sutter, J.; Miller, J. T.; Meyer, K.; Cantat, T.; Diaconescu, P. L. Redox control of a polymerization catalyst by changing the oxidation state of the metal center. Chem. Commun. 2011, 47, 9897−9899. (53) Sauer, A.; Buffet, J.-C.; Spaniol, T. P.; Nagae, H.; Mashima, K.; Okuda, J. Switching the Lactide Polymerization Activity of a Cerium Complex by Redox Reactions. ChemCatChem 2013, 5, 1088−1091. (54) Biernesser, A. B.; Li, B.; Byers, J. A. Redox-Controlled Polymerization of Lactide Catalyzed by Bis(imino)pyridine Iron Bis(alkoxide) Complexes. J. Am. Chem. Soc. 2013, 135, 16553−16560. (55) Fang, Y.-Y.; Gong, W.-J.; Shang, X.-J.; Li, H.-X.; Gao, J.; Lang, J.-P. Synthesis and structure of a ferric complex of 2,6-di(1H-pyrazol-3yl)pyridine and its excellent performance in the redox-controlled living ring-opening polymerization of ε-caprolactone. Dalton Trans. 2014, 43, 8282−8289. (56) Yoon, H. J.; Kuwabara, J.; Kim, J.-H.; Mirkin, C. A. Allosteric Supramolecular Triple-Layer Catalysts. Science 2010, 330, 66−69. (57) Coulembier, O.; Moins, S.; Todd, R.; Dubois, P. External and Reversible CO2 Regulation of Ring-Opening Polymerizations Based on a Primary Alcohol Propagating Species. Macromolecules 2014, 47, 486− 491. (58) Bianchi, J. P.; Price, F. P.; Zimm, B. H. “Monodisperse” Polystyrene. J. Polym. Sci. 1957, 25, 27−38. 1990

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

(78) Alfredo, N. V.; Jalapa, N. E.; Morales, S. L.; Ryabov, A. D.; Le Lagadec, R.; Alexandrova, L. Light-Driven Living/Controlled Radical Polymerization of Hydrophobic Monomers Catalyzed by Ruthenium(II) Metalacycles. Macromolecules 2012, 45, 8135−8146. (79) Xu, J.; Jung, K.; Boyer, C. Oxygen Tolerance Study of Photoinduced Electron Transfer - Reversible Addition-Fragmentation Chain Transfer (PET-RAFT) Polymerization Mediated by Ru(bpy)3Cl2. Macromolecules 2014, 47, 4217−4229. (80) Xu, J.; Jung, K.; Corrigan, N. A.; Boyer, C. Aqueous photinduced living/controlled polymerization: tailoring for bioconjugation. Chem. Sci. 2014, 5, 3568−3575. (81) Xu, J.; Boyer, C. Visible Light Photocatalytic Thiol-Ene Reaction: An Elegant Approach for Fast Polymer Postfunctionalization and StepGrowth Polymerization. Macromolecules 2015, 48, 520−529. (82) Fors, B. P.; Hawker, C. J. Control of a Living Radical Polymerization of Methacrylates by Light. Angew. Chem., Int. Ed. 2012, 51, 8850−8853. (83) Treat, N. J.; Fors, B. P.; Kramer, J. W.; Christianson, M.; Chiu, C.Y.; Read de Alaniz, J.; Hawker, C. J. Controlled Radical Polymerization of Acrylates Regulated by Visible Light. ACS Macro Lett. 2014, 3, 580− 584. (84) Xu, J.; Jung, K.; Atme, A.; Shanmugam, S.; Boyer, C. A Robust and Versatile Photoinduced Living Polymerization of Conjugated and Unconjugated Monomers and Its Oxygen Tolerance. J. Am. Chem. Soc. 2014, 136, 5508−5519. (85) Shanmugam, S.; Xu, J.; Boyer, C. Photoinduced Electron Transfer-Reversible Addition-Fragmentation Chain Transfer (PETRAFT) Polymerization of Vinyl Acetate and N-Vinylpyrrolidinone: Kinetic and Oxygen Tolerance Study. Macromolecules 2014, 47, 4930− 4942. (86) Fu, C.; Xu, J.; Tao, L.; Boyer, C. Combining Enzymatic Monomer Transformation with Photoinduced Electron Transfer-Reversible Addition-Fragmentation Chain Transfer for the Synthesis of Complex Multiblock Copolymers. ACS Macro Lett. 2014, 3, 633−638. (87) Shanmugam, S.; Boyer, C. Stereo-, Temporal and Chemical Control through Photoactivation of Living Radical Polymerization: Synthesis of Block and Gradient Copolymers. J. Am. Chem. Soc. 2015, 137, 9988−9999. (88) Liu, Q.; Liu, L.; Ma, Y.; Zhao, C.; Yang, W. Visible Light-Induced Controlled Radical Polymerization of Methacrylates with Perfluoroalkyl Iodide as the Initiator in Conjugation with a Photoredox Catalyst fac[Ir(ppy)3]. J. Polym. Sci., Part A: Polym. Chem. 2014, 52, 3283−3291. (89) Telitel, S.; Dumur, F.; Telitel, S.; Soppera, O.; Lepeltier, M.; Guillaneuf, Y.; Poly, J.; Morlet-Savary, F.; Fioux, P.; Fouassier, J.-P.; Gigmes, D.; Lalevée, J. Photoredox catalysis using a new iridium complex as an efficient toolbox for radical, cationic and controlled polymerizations under soft blue to green lights. Polym. Chem. 2015, 6, 613−624. (90) Zhang, J.; Li, A.; Liu, H.; Yang, D.; Liu, J. Well-Controlled RAFT Polymerization Initiated by Recyclable Surface-Modified Nb(OH)5 Nanoparticles Under Visible Light Irradiation. J. Polym. Sci., Part A: Polym. Chem. 2014, 52, 2715−2724. (91) Cao, Y.; Xu, Y.; Zhang, J.; Yang, D.; Liu, J. Well-controlled atom transfer radical polymerizations of acrylates using recyclable niobium complex nanoparticle as photocatalyst under visible light irradiation. Polymer 2015, 61, 198−203. (92) Treat, N. J.; Sprafke, H.; Kramer, J. W.; Clark, P. G.; Barton, B. E.; Read de Alaniz, J.; Fors, B. P.; Hawker, C. J. Metal-Free Atom Transfer Radical Polymerization. J. Am. Chem. Soc. 2014, 136, 16096−16101. (93) Miyake, G. M.; Theriot, J. C. Perylene as an Organic Photocatalyst for the Radical Polymerization of Functionalized Vinyl Monomers through Oxidative Quenching with Alkyl Bromides and Visible Light. Macromolecules 2014, 47, 8255−8261. (94) Chen, M.; MacLeod, M. J.; Johnson, J. A. Visible-LightControlled Living Radical Polymerization from a Trithiocarbonate Iniferter Mediated by an Organic Photoredox Catalyst. ACS Macro Lett. 2015, 4, 566−569. (95) Xu, J.; Shanmugam, S.; Duong, H. T.; Boyer, C. Organophotocatalysts for photoinduced electron transfer-reversible addition-

(59) Konkolewicz, D.; Wang, Y.; Zhong, M.; Krys, P.; Isse, A. A.; Gennaro, A.; Matyjaszewski, K. Reversible-Deactivation Radical Polymerization in the Presence of Metallic Copper. A Critical Assessment of the SARA ATRP and SET-LRP Mechanisms. Macromolecules 2013, 46, 8749−8772. (60) Kwak, Y.; Matyjaszewski, K. Photoirradiated Atom Transfer Radical Polymerization with an Alkyl Dithiocarbamate at Ambient Temperature. Macromolecules 2010, 43, 5180−5183. (61) Tasdelen, M. A.; Uygun, M.; Yagci, Y. Photoinduced Controlled Radical Polymerization in Methanol. Macromol. Chem. Phys. 2010, 211, 2271−2275. (62) Murtezi, E.; Yagci, Y. Simultaneous Photoinduced ATRP and CuAAC Reactions for the Synthesis of Block Copolymers. Macromol. Rapid Commun. 2014, 35, 1782−1787. (63) Doran, S.; Yagci, Y. Graft polymer growth using tandem photoinduced photoinitiator-free CuAAC/ATRP. Polym. Chem. 2015, 6, 946−952. (64) Ciftci, M.; Tasdelen, M. A.; Li, W.; Matyjaszewski, K.; Yagci, Y. Photoinitiated ATRP in Inverse Microemulsion. Macromolecules 2013, 46, 9537−9543. (65) Dadashi-Silab, S.; Tasdelen, M. A.; Asiri, A. M.; Khan, S. B.; Yagci, Y. Photoinduced Atom Transfer Radical Polymerization Using Semiconductor Nanoparticles. Macromol. Rapid Commun. 2014, 35, 454−459. (66) Dadashi-Silab, S.; Tasdelen, M. A.; Kiskan, B.; Wang, X.; Antonietti, M.; Yagci, Y. Photochemically Mediated Atom Transfer Radical Polymerization Using Polymeric Semiconductor Mesoporous Graphitic Carbon Nitride. Macromol. Chem. Phys. 2014, 215, 675−681. (67) Anastasaki, A.; Nikolaou, V.; Zhang, Q.; Burns, J.; Samanta, S. R.; Waldron, C.; Haddleton, A. L.; McHale, R.; Fox, D.; Percec, V.; Wilson, P.; Haddleton, D. M. Copper(II)/Tertiary Amine Synergy in Photoinduced Living Radical Polymerization: Accelerated Synthesis of ωFunctional and α,ω-Heterofunctional Poly(acrylates). J. Am. Chem. Soc. 2014, 136, 1141−1149. (68) Anastasaki, A.; Nikolaou, V.; Brandford-Adams, F.; Nurumbetov, G.; Zhang, Q.; Clarkson, G. J.; Fox, D. J.; Wilson, P.; Kempe, K.; Haddleton, D. M. Photo-induced living radical polymerization of acrylates utilizing a discrete copper(II)-formate complex. Chem. Commun. 2015, 51, 5626−5629. (69) Ciftci, M.; Tasdelen, M. A.; Yagci, Y. Sunlight induced atom transfer radical polymerization by using dimanganese decacarbonyl. Polym. Chem. 2014, 5, 600−606. (70) Mosnácě k, J.; Ilčíková, M. Photochemically Mediated Atom Transfer Radical Polymerization of Methyl Methacrylate Using ppm Amounts of Catalyst. Macromolecules 2012, 45, 5859−5965. (71) Mosnácě k, J.; Eckstein-Andicsová, A.; Borská, K. Ligand effect and oxygen tolerance studies in photochemically induced copper mediated reversible deactivation radical polymerization of methyl methacrylate in dimethyl sulfoxide. Polym. Chem. 2015, 6, 2523−2530. (72) Konkolewicz, D.; Schröder, K.; Buback, J.; Bernhard, S.; Matyjaszewski, K. Visible Light and Sunlight Photoinduced ATRP with ppm of Cu Catalyst. ACS Macro Lett. 2012, 1, 1219−1223. (73) Zhang, T.; Chen, T.; Amin, I.; Jordan, R. ATRP with a light switch: photoinduced ATRP using a household fluorescent lamp. Polym. Chem. 2014, 5, 4790−4796. (74) Magenau, A. J. D.; Strandwitz, N. C.; Gennaro, A.; Matyjaszewski, K. Electrochemically Mediated Atom Transfer Radical Polymerization. Science 2011, 332, 81−84. (75) Li, B.; Yu, B.; Huck, W. T. S.; Liu, W.; Zhou, F. Electrochemically Mediated Atom Transfer Radical Polymerization on Nonconducting Substrates: Controlled Brush Growth through Catalyst Diffusion. J. Am. Chem. Soc. 2013, 135, 1708−1710. (76) Iwai, K.; Uesugi, M.; Takemura, F. Tris(2,2′Bipyridine)Ruthenium(II)-Sensitized Photopolymerization of Acrylamide. Polym. J. 1985, 17, 1005−1011. (77) Zhang, G.; Song, I. Y.; Ahn, K. H.; Park, T.; Choi, W. Free Radical Polymerization Initiated and Controlled by Visible Light Photocatalysis at Ambient Temperature. Macromolecules 2011, 44, 7594−7599. 1991

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992

Chemical Reviews

Review

fragmentation chain transfer (PET-RAFT) polymerization. Polym. Chem. 2015, 6, 5615−5624. (96) Shanmugam, S.; Xu, J.; Boyer, C. Utilizing the electron transfer mechanism of chlorophyll a under light for controlled radical polymerization. Chem. Sci. 2015, 6, 1341−1349. (97) Shanmugam, S.; Xu, J.; Boyer, C. Exploiting Metalloporphyrins for Selective Living Radical Polymerization Tunable over Visible Wavelengths. J. Am. Chem. Soc. 2015, 137, 9174−9185. (98) For the application of metalloporphyrins in immortal polymerizations, see: Inoue, S. Immortal Polymerization: The Outset, Development, and Application. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 2861−2871. (99) Koumura, K.; Satoh, K.; Kamigaito, M. Manganese-Based Controlled/Living Radical Polymerization of Vinyl Acetate, Methyl Acrylate, and Styrene: Highly Active, Versatile, and Photoresponsive Systems. Macromolecules 2008, 41, 7359−7367. (100) Miao, X.; Li, J.; Zhang, Z.; Cheng, Z.; Zhang, W.; Zhu, J.; Zhu, X. Dimanganese decacarbonyl/2-cyanoprop-2-yl-1-dithionaphthalate: toward sunlight induced RAFT polymerization of MMA. Polym. Chem. 2014, 5, 4641−4648. (101) Yamago, S.; Ukai, Y.; Matsumoto, A.; Nakamura, Y. Organotellurium-Mediated Controlled/Living Radical Polymerization Initiated by Direct C−Te Bond Photolysis. J. Am. Chem. Soc. 2009, 131, 2100− 2101. (102) Zhao, Y.; Yu, M.; Fu, X. Photo-cleavage of the cobalt−carbon bond: visible light-induced living radical polymerization mediated by organo-cobalt porphyrins. Chem. Commun. 2013, 49, 5186−5188. (103) Zhao, Y.; Yu, M.; Zhang, S.; Liu, Y.; Fu, X. Visible Light Induced Living/Controlled Radical Polymerization of Acrylates Catalyzed by Cobalt Porphyrins. Macromolecules 2014, 47, 6238−6245. (104) Shi, Y.; Liu, G.; Gao, H.; Lu, L.; Cai, Y. Effect of Mild Visible Light on Rapid Aqueous RAFT Polymerization of Water-Soluble Acrylic Monomers at Ambient Temperature: Initiation and Activation. Macromolecules 2009, 42, 3917−3926. (105) Tong, J.; Shi, Y.; Liu, G.; Huang, T.; Xu, N.; Zhu, Z.; Cai, Y. Visible Light Mediated Fast Iterative RAFT Synthesis of Amino-Based Reactive Copolymers in Water at 20 °C. Macromol. Rapid Commun. 2013, 34, 1827−1832. (106) Ohtsuki, A.; Goto, A.; Kaji, H. Visible-Light-Induced Reversible Complexation Mediated Living Radical Polymerization of Methacrylates with Organic Catalysts. Macromolecules 2013, 46, 96−102. (107) Ohtsuki, A.; Lei, L.; Tanishima, M.; Goto, A.; Kaji, H. Photocontrolled Organocatalyzed Living Radical Polymerization Feasible over a Wide Range of Wavelengths. J. Am. Chem. Soc. 2015, 137, 5610−5617. (108) Zhou, H.; Johnson, J. A. Photo-controlled Growth of Telechelic Polymers and End-linked Polymer Gels. Angew. Chem., Int. Ed. 2013, 52, 2235−2238. (109) Wolpers, A.; Vana, P. UV Light as External Switch and Boost of Molar-Mass Control in Iodine-Mediated Polymerization. Macromolecules 2014, 47, 954−963. (110) Murata, K.; Saito, K.; Kikuchi, S.; Akita, M.; Inagaki, A. Visiblelight-controlled homo- and copolymerization of styrenes by a bichromophoric Ir-Pd catalyst. Chem. Commun. 2015, 51, 5717−5720. (111) Frisch, H.; Nie, Y.; Raunser, S.; Besenius, P. pH-Regulated Selectivity in Supramolecular Polymerizations: Switching between Coand Homopolymers. Chem. - Eur. J. 2015, 21, 3304−3309.

1992

DOI: 10.1021/acs.chemrev.5b00426 Chem. Rev. 2016, 116, 1969−1992