Keywords

1 Introduction

Living anionic polymerization provides a versatile method for the synthesis of macromolecules having a low degree of compositional heterogeneity [1, 2]. The term living refers to the ability of the growing polymer anions to propagate without termination or chain transfer reactions, which was elegantly demonstrated by Michael Szwarc in 1956 [3–5]. Since Szwarc’s discovery, anionic polymerization has been exploited to achieve the controlled polymerization of numerous suitable monomers, as well as the synthesis of linear block copolymers (diblock, triblock, etc.) and various types of branched polymers (stars, combs, and dendrimers) [1, 2, 5–10]. Incorporation of two or more monomers to produce linear block and branched copolymer systems, such as miktoarm stars and graft copolymers, has resulted in a broad range of elegant multiphase tailored materials based on self-assembly of incompatible segments [11]. Polymers exhibiting these complicated structures have sparked the interest of theoreticians, physicists, and synthetic chemists alike, not only because of their exploitation for new applications but for being both processable and recyclable [12].

Well-defined branched structures have steadily gained attention because of their unique properties which can be tuned through chemical design; therefore, these materials can address numerous applications ranging from thermoplastic elastomers and high-impact plastics to pressure-sensitive adhesives, additives, and foams [6]. In particular, graft copolymers with desired functional groups, chemical compositions, and a large grafting density have been of interest for numerous biological and nanoscience applications [13–18]. Over the past quarter century, the synthesis of model branched structures has expanded to include not just anionic polymerization but a variety of living/controlled polymerizations, as well as approaches incorporating a combination of techniques. These developments have been promoted by advancements in anionic, cationic, and radical polymerization methods including: ring-opening polymerization (ROP), ring-opening metathesis polymerization (ROMP), atom transfer radical polymerization (ATRP), single electron transfer living radical polymerization (SET-LRP), reversible addition-fragmentation chain transfer (RAFT) polymerization, and nitroxide-mediated polymerization (NMP) [7, 8, 13]. Although anionic polymerization techniques are limited to a rather small range of monomers, this method offers the maximum control over the polymerization with the absence of chain transfer or other termination reactions, whereas controlled radical methods do suffer from termination events resulting in broader polydispersity indices (PDI – the ratio of weight-average to number-average molecular weights), incomplete monomer conversions (p < 0.8), and decreased grafting efficiencies (f < 0.75) [8, 19–22]. Strong synthetic control is not only significant to synthetic polymer chemists, but the availability of well-defined precisely tailored polymers is critical to polymer physicists and engineers for use in gaining a fundamental understanding of the correlation between polymer architecture, molecular composition, and physical properties in order to tailor polymers for specific applications [13].

2 General Aspects of Graft Copolymer Synthesis

Comb and graft copolymer architectures consist of a linear polymeric backbone having one or more polymer side chains attached by covalent bonds [1, 6]. Comb structures are the simplest form of these branched architectures, where both the main chain and branches have the same chemical composition. In contrast, graft copolymers are comprised of a backbone and side chains that differ in chemical composition [9, 23]. The structures of both comb and graft copolymers are defined by (1) the molecular weight of the main chain, (2) the molecular weight of the graft chain, and (3) the distance between the graft chains [24] (Scheme 1).

Scheme 1
scheme 1

Three structural factors of graft copolymers [24]

Full size image

Optimum control of the polymerization provides a basis for the precision synthesis of well-defined graft copolymers. Ideally, these comb and graft copolymer materials would consist of monodisperse side chains bound to a monodisperse main chain, with the number of branches, branch spacing, and branch point functionality all being precisely controlled [25]. This has been most nearly achieved through anionic polymerization, providing an array of branched structures, comprised of a variety of monomers, and exhibiting narrow PDIs with tight control over branch junction functionality and placement, as discussed below. However, most anionic graft copolymer syntheses reported to date, because of their mechanisms, yield a controlled average number of graft branches per molecule and random spacing distribution of graft branches along the backbone [1].

Anionic polymerization is well suited for the synthesis of controlled comb and grafted architectures due to the absence of termination and other transfer reactions that results in an increase of heterogeneity within the polymerization [2, 26–33]. Various other polymerization methods including conventional free radical, living free radical, and cationic polymerizations have been used to synthesize comb and graft polymers, as discussed in detail in prior works [8, 33–35]. It is important to note that the primary advantages of using anionic polymerization approach for the construction of comb and other branched structures include, but not limited to, high grafting efficiency, control of the backbone and side chain molecular weights and polydispersities, and in some cases control of branch point functionality and spacing. However, this methodology is often restricted because of enduring experimental techniques and characterization shortcomings. The following sections provide a general overview of anionic polymerization methods for the synthesis of graft copolymers. It is important to note that the number of branch sites can be controlled through stoichiometric measures and is randomly distributed along the backbone unless specified.

There are three general methods for the synthesis of grafted polymers (Scheme 2): (a) “grafting onto,” where the backbone polymer chain contains heterogeneously placed functional groups, X, that will react with another macromolecule with a chain end antagonistic reactive functional group, Y, a living anion in the case of anionic polymerization; (b) “grafting from,” where the active sites are generated along the polymeric backbone, giving way to a pseudo-multifunctional macroinitiator, to be used in the initiation of the second monomer; and (c) “grafting through,” in which a living polymer chain is end capped with an unsaturated monomeric head unit, forming a macromonomer that will undergo further homo- or copolymerization during the backbone construction [1]. We will review these three methods for comb polymer and graft copolymer synthesis, citing classic early work and also highlighting recent examples over the past decade. This will be followed by a discussion of the most recent specialized methods for synthesis of combs and graft copolymers, which offer superior control over macromolecular architecture. There is no effort made here to cite all examples, especially those from older literature. Readers are referred to earlier reviews on the subject [1, 2, 8, 10, 24, 36, 37].

Scheme 2
scheme 2

Three basic strategies for synthesis of comb polymers and graft copolymers [29]

Full size image

2.1 Grafting Onto

The grafting onto method involves the nucleophilic attack of the living polymer side chains along the main chain at suitable electrophilic sites, with anhydrides, esters, pyridine, or benzylic halide groups among the most commonly utilized [8, 10, 25]. Branching sites along the backbone can be generated by post-polymerization modification or by copolymerization with monomer(s) bearing the desired functional group. Under appropriate reaction conditions, the backbone and side chains undergo a coupling reaction, covalently bonding the two, resulting in the final comb or graft copolymer architecture. A key advantage to this method is that before the coupling reaction, both the living anionic side chains and polymerized backbone may be isolated, allowing for independent characterization. Measuring the molecular weights of the grafted product and the homo- or copolymer backbone and side chain precursors allows the number of branches and the grafting efficiency to be more accurately determined. Additionally, in cases where both the backbone and side chains are polymerized by anionic methods, the molecular weight, polydispersity, and the chemical composition of each can be controlled, yielding near homogeneous comb- and graft-type products [10].

The most common procedure to synthesize comb or graft structures by a grafting onto approach utilizes chloromethylation of polystyrene (PS) (Scheme 3) [38]. Using this method and living poly(ethylene oxide) (PEO) oxyanions made by anionic polymerization, Rempp and coworkers synthesized PS-g-PEO graft copolymers [39]. However, reaction of many polymeric carbanions with chloromethyl groups results in undesirable metal-halogen exchange, altering the functionality of the branched polymer [40–42]. Conversion of chloromethyl groups of the modified PS into chlorosilyl moieties, established by Rahlwes and coworkers, results in quantitative reaction with poly(isoprenyllithium) to make PS-g-PI, where PI is polyisoprene [43]. Additional PS graft copolymers containing poly(2-vinylpyridine) (P2VP), poly(4-vinylpyridine) (P4VP), poly(methyl methacrylate) (PMMA), and poly(tert-butylmethacrylate) (PtBuMA) side chains were synthesized by partial chloromethylation or bromomethylation of the anionically prepared PS main chain followed by reaction with the living anions of the side chains [39, 44–47]. Hydrolysis of the tert-butyl group of PtBuMA leads to PS-g-PMA, where PMA is poly(methacrylic acid) [48].

Scheme 3
scheme 3

Chloromethylation of PS [38]

Full size image

All these linking reactions occur through nucleophilic attack of the living polyanion upon the backbone. A successful reaction often requires reducing the reactivity of the anions to avoid side reactions. Living polycarbanions may be end capped with 1,1-diphenylethylene (DPE) [49] or ethylene oxide for this purpose. DPE end capping of PS anions was employed by Fernyhough et al. [50], followed by reaction with a chloromethylated PS backbone and sulfonation to produce comb-branched sodium poly(styrenesulfonate)s (NaPSS).

Using a modification of this strategy, Wilhelm and coworkers recently reported the synthesis and characterization of well-defined model comb polymers of poly(4-methylstyrene) (P4MS) having very low, but well-controlled, degrees of branching [51]. The backbone and side chains of P4MS were made by living anionic polymerization. The backbone was then treated with N-bromosuccinimide (NBS) to partially brominate the methyl groups of the backbone polymer, and the living P4MS side chains were end capped with DPE prior to linking. For a degree of bromination, >2 mol% side reactions leading to cross-linking occurred, likely a result of metal-halogen exchange reactions.

Kawahawa et al. reacted a partially brominated polypropylene (PP) backbone, made by a metallocene-catalyzed copolymerization of propylene with 11-bromo-1-undecene, with PSLi to create PP-g-PS [52]. The obtained graft copolymers have potential as compatibilizers for blends of PP and PS. Lin and coworkers prepared an iodinated PBd backbone and demonstrated the synthesis of comb PBd and various graft copolymers by reaction with DPE end-capped living polyanions of various types, including various block copolymers to create complex copolymer architectures including more than one type of grafted side chain [53]. Tang et al. synthesized various comb and graft structures with PI as the backbone by epoxidation of anionically synthesized PI using H2O2/HCOOH, followed by reaction with PI, PS, PS-b-PI, and PI-b-PS [54].

More complex architectures have been synthesized by modification of this grafting onto strategy. Graft copolymers with “V-shaped” and “Y-shaped” side chains were synthesized by a combination of controlled radical polymerization and living anionic polymerization (Scheme 4) [55]. The backbone was prepared by TEMPO-mediated copolymerization of styrene and vinylbenzyl chloride. The V-shaped branches were made by reaction of a PS macromonomer, having a DPE end group, with PSLi or PILi, followed by reaction with the benzyl chloride-functionalized PS backbone. The Y-shaped branches were incorporated in a similar manner after further polymerization initiated by the “V-shaped” polyanions.

Scheme 4
scheme 4

Synthesis of graft (co)polymers with V- and Y-shaped branches [55]

Full size image

Borsali and Deffieux and coworkers synthesized PS-b-PI comb copolymers by the selective coupling of PS anions with a poly(chloroethylvinylether) backbone block, followed by reaction of PI anions with chlorobutyl functionality introduced into the second backbone block [56]. The same group synthesized comblike copolymers having randomly distributed PS and PI grafts attached to a poly(chloroethylvinylether) backbone by a grafting onto strategy [57].

Hirao et al. prepared poly(m-tert-butyldimethylsiloxymethylstyrene)s with narrow PDIs by living anionic polymerization, transformed them into poly(m-chloromethylstyrene)s by reaction with BCl3, and reacted this functionalized backbone with living anionic polymers of styrene, isoprene, and 2-vinylpyridine to synthesize comb polymers and graft copolymers having one branch on each repeating unit (macromolecular bottlebrush) [58, 59]. Attempts to produce polymers carrying two branches per repeat unit were carried out using modifications of this approach [59].

Polybutadiene (or alternatively polyisoprene) can be chlorosilylated for preparation of PBd-g-PS graft copolymers and PBd-g-PBd combs, where PBd is polybutadiene [60–62]. The synthesis of PBd-g-PS begins with the anionic polymerization of butadiene in benzene, resulting in a linear PBd backbone with >90 % 1,4-addition. Post-polymerization hydrosilylation using (CH3)2SiHCl creates chlorosilane groups at pendant double bonds of the 1,2-polybutadiene units. The reaction with living PS anions or PBd anions gave PBd-g-PS or PBd-g-PBd, respectively, with random branch placement (Scheme 5). Increased functionality of the branching sites can be introduced through the use of multifunctional Si-Cl coupling agents during the hydrosilylation step [10, 63, 64]. The same synthetic scheme produces comb-branched polyethylenes when catalytic hydrogenation is carried out on PBd comb homopolymers [61, 62].

Scheme 5
scheme 5

Hydrosilylation of PBd backbone as a route to grafted polymers and copolymers [60]

Full size image

The anionic copolymerization of the difunctional monomer 4-(vinylphenyl)-1-butene with styrene at −40 °C in toluene/THF to yield PS-g-PS, PS-g-PI, and PS-g-PMMA has also been demonstrated [65]. The experimental conditions permit the styrenic double bond to be more reactive, giving a linear backbone with randomly spaced alkene pendent groups. The alkene functionality allows the introduction of Si-Cl groups via hydrosilylation at the olefinic double bonds and when introduced to suitable living polyanions (PS, PI, PMMA) produces well-defined comb polymers or graft copolymers.

This same group described the synthesis of amphiphilic graft copolymers composed of hydrophilic backbones and hydrophobic PS side chains with various acrylic monomers [66, 67]. The synthesis begins with the anionic copolymerization of the desired methacrylic backbone monomer and glycidyl methacrylate (GMA). Living polystyryllithium (PSLi) is then added, which undergoes a rapid reaction between the active PS chain ends and pendent epoxy groups of GMA, forming the desired final graft copolymer composition. The amount of GMA in the copolymer backbone controls the number of available grafting sites along the backbone. After reaction with living side chains, all unreacted epoxy groups are neutralized with 1,1-diphenylhexyllithium [66, 67].

A variety of other grafted homopolymers and copolymers have been prepared by reacting living polyanions with suitable functionalized polymer backbones including poly(N-vinylcarbazole) [68], cellulose [69, 70], and maleic anhydride copolymers [71].

Block-graft copolymers are block copolymers having one or more linear block and one or more blocks which are graft copolymers [72]. Poly[styrene-b-(4-vinylphenyl)dimethylvinylsilane)-g-polyisoprene]-b-polystyrene block-graft copolymers were synthesized by Se et al. [73]. (4-Vinylphenyl)dimethylvinylsilane was anionically polymerized using cumylcesium in THF at −78 °C, followed by addition of styrene to create a PVS-b-PS diblock copolymer backbone. Living PI anions were reacted with the PVS functional groups to create the block-graft architecture. Block-graft copolymers having a PS backbone with PI side chains attached to only a portion of the backbone were made by a grafting onto strategy [74]. The backbone block copolymer was made by using TEMPO-mediated radical copolymerization of styrene and vinylbenzyl chloride to grow the first block, followed by addition of styrene to grow the second block. Living PI anions were reacted with benzyl chloride groups in the first block in order to attach the side chains.

2.2 Grafting From

The grafting from method employs creation of active sites along the polymer backbone, which serve to initiate the polymerization of the monomer that will create the side chains. A disadvantage of this approach, as compared to the grafting onto method, is that the side chains cannot usually be isolated and characterized independently, making it more difficult to ascertain side chain length and grafting density. As with the grafting onto strategy, if the backbone and side chains are both made via a controlled/living polymerization process, they can be of narrow PDI. Branch placement is generally random.

Grafting from by anionic polymerization is most often accomplished through acid-base chemistry [25]. Acidic hydrogens on amide, alcohol, or phenol groups can be removed by treatment with tert-BuOK; subsequently, the anionic polymerization of ethylene oxide may then proceed [75]. Acidic hydrogens, adjacent to a carbonyl group, can be removed with lithium diisopropylamide (LDA) and have been shown to be well suited for the polymerization of methacrylate monomers [76].

Metallation by organometallic compounds such as s-BuLi, in the presence of a strong chelating agent, i.e., N,N,N’N’-tetramethylethylenediamine (TMEDA), has been shown to create main chain active sites of allylic, benzylic, and aromatic C-H bonds (Scheme 6). Several groups demonstrated this technique by producing various poly(diene-g-styrene)s [77–82]. Hadjichristidis and Roovers synthesized model poly(isoprene-g-styrene) via lithiation of PI in the presence of TMEDA and improved this grafting chemistry by performing all manipulations under high vacuum in sealed vessels followed by fractionation to isolate the desired products [78].

Scheme 6
scheme 6

Metallation of a polydiene backbone, followed by monomer addition to create a graft copolymer by the “grafting from” mechanism [1]

Full size image

An important advancement in the metallation technique by the grafting from approach was the introduction of metallation via a superbase. Lochmann et al. reported the one-pot metallation of a parent PS backbone by a superbase, prepared from 3-lithiomethylheptane and excess potassium alkoxide, followed by the reaction with an electrophile to produce randomly substituted polystyrene containing a variety of side chain pendant groups [83]. Using a less electronegative potassium counterion enhances the reactivity by promoting the favorable association of Li and O, generating potassium-contained substituents along the PS backbone. The superbase reagent provides for rapid deprotonation, at room temperature in cyclohexane, of modestly acidic substances, and the high reactivity of the organopotassium sites is capable of reacting with a large spectrum of electrophiles [83–85]. Therefore, graft copolymers may be synthesized by polymerization or copolymerization of a suitable monomer (e.g., 4-methylstyrene), metallation, and addition of an anionically polymerizable monomer (Scheme 7) [84, 85].

Scheme 7
scheme 7

Synthesis of graft copolymers through metallation of poly(styrene-co-methylstyrene) [85]

Full size image

Employing this general strategy, PP-g-PS was recently synthesized by Li et al. [86]. They used Ziegler-Natta copolymerization of propylene and p-allyltoluene to prepare the backbone, followed by metallation of the benzylic sites using n-BuLi/tert-BuOK and addition of styrene monomer. Polypropylene-g-poly(ethylene-co-1-butene) with a well-defined long-chain branched structure was prepared by a similar strategy (metallocene-initiated copolymerization of propylene and 4-(3-butenyl)toluene, followed by metallation of benzylic sites using sec-BuLi/TMEDA, addition of butadiene monomer to grow PBd side chain, and hydrogenation) (Scheme 8) [87].

Scheme 8
scheme 8

Synthesis of PP-g-EBR copolymers [87]

Full size image

Se et al. [88, 89] prepared di- and triblock copolymers of styrene and p-tert-butoxystyrene to prepare a polymer backbone, followed by acid hydrolysis to convert the butoxy groups to hydroxyl groups and activation using cumyl potassium or DPE potassium, and the subsequent addition of ethylene oxide yields the block-graft copolymers. Pispas and coworkers used anionic copolymerization of styrene and p-tert-butoxystyrene, which when deprotected was used to polymerize ethylene oxide in a ring-opening anionic polymerization using phosphazene base (t-BuP4) (Scheme 9) [90]. The same team synthesized thermo-responsive brush copolymers with poly(propylene oxide-r-ethylene oxide) side chains via the same strategy employing a metal-free anionic polymerization grafting from strategy [91].

Scheme 9
scheme 9

Synthetic scheme for the preparation of polystyrene-b-poly(p-hydroxystyrene-g-ethylene oxide) [90]

Full size image

In another approach, PMMA-g-poly(β-butyrolactone) copolymers have been synthesized [92]. This method involves treating the PMMA backbone with 18-crown-6 and potassium hydroxide to create the macroinitiator. This reaction was determined to have a high grafting efficiency with facile control of the density of grafting sites. Graft copolymers of PMMA-g-polyamide 6 were synthesized by using N-carbamated caprolactam pendants as macroactivators and sodium caprolactamate as a catalyst for the anionic polymerization of caprolactam [93]. This group used a similar approach for synthesis of various graft copolymers with “rigid” backbones via anionic ring-opening polymerization of caprolactam (Scheme 10) [94, 95].

Scheme 10
scheme 10

Mechanism of graft copolymers of (a) St, (b) MMA, (c) NPMI, (d) ACCL, (e) macroactivator, and (f) graft copolymer [94]

Full size image

Graft copolymers having aromatic polyimide backbones and caprolactam side chains were synthesized by anionic polymerization of caprolactam initiated by imide rings bearing several suitable functional groups [96]. PS-g-polyamide 6 has been synthesized by anionic polymerization of caprolactam from an isocyanate-bearing PS backbone [97]. Amphiphilic polycaprolactam-g-PMMA and polycaprolactam-g-poly(L-lysine) copolymers have been synthesized by metallation of a PCL backbone using LDA/THF, followed by addition of suitable monomers [98, 99]. These degradable copolymers are of interest in biomedical and pharmaceutical applications. Degradable amphiphilic polycaprolactam-g-P4VP and polycaprolactam-g-poly(dimethylaminoethyl methacrylate) copolymers have been reported by the same group [100]. More recently, Schlaad reported the anionic graft copolymerization of ethylene oxide directly from poly(N-isopropyl acrylamide) via activation of the PNIPAM using phosphazene base [101].

The grafting from strategy via living anionic synthetic procedures has two major disadvantages: (1) knowledge of the precursor polymer chain is generally limited, as noted above, to the backbone, with the isolation and characterization of the branches generally involving selective chemical decomposition, and (2) ionic copolymerization can lead to poor solubility, which results in poor control of the polymerization [9, 10, 25, 65]. However, this methodology is considered particularly attractive for use in controlled radical polymerization techniques since there is a low concentration of instantaneous propagating species present, limiting coupling and other termination events, and the continuous growth of side chains effectively relieves steric effects [13].

2.3 Grafting Through (The Conventional Macromonomer Approach)

The grafting through method relies on the formation and polymerization of macromonomers, which are oligo- or polymeric chains characterized by a polymerizable head group. Following this methodology, the side chains are first covalently bonded to a polymerizable moiety, forming the macromonomer. When homopolymerized, the macromonomer results in molecular brushes, but it is more commonly copolymerized to produce comb polymers and graft copolymers. The macromonomer approach requires consideration of important synthetic factors but offers access to reasonably well-defined grafted architectures, having well-defined side chains and backbones, more easily than other grafting methods based upon anionic polymerization. The most important consideration is the reactivity disparity of the macromonomer (M 1) and comonomer (M 2), typically expressed by reactivity ratios r 1 and r 2, described by the Mayo-Lewis copolymerization equation [102]:

$$ \mathrm{d}\left[{M}_2\right]/\mathrm{d}\left[{M}_1\right] = \left(1 + {r}_2\left[{M}_2\right]/\left[{M}_1\right]\right)\ /\ \left(1 + {r}_1\left[{M}_1\right]/\left[{M}_2\right]\right). $$
(1)

Generally, ionic mechanisms exhibit a greater discrepancy between r 1 and r 2, as compared to free radical copolymerization, resulting in limited control of branch placement along the backbone. Additional factors of inherent incompatibility between the macromonomer and growing polymer chains, fluctuations in concentrations of the two compounds, and phase separation due to the formation of the copolymer can lead to greater compositional and molecular weight heterogeneity of the product [9, 10, 25].

The formation of PS and PMMA macromonomers has been described in numerous publications [103–108]. An example of the PS macromonomer synthesis begins with the homopolymerization of styrene initiated by sec-BuLi. The living polymer is then end capped with ethylene oxide to increase chain end flexibility and reactivity before reacting with methacryloyl chloride to yield a methacrylate polymerizable end group or benzyl bromide (or chloride) for a styrenic polymerizable end group. Living PS has also been end capped with 1,1-dipheylethylene and subsequently reacted with vinyl benzyl bromide (or chloride). Similarly, PMMA macromonomers were synthesized, end capped, and copolymerized with MMA or styrene to yield PMMA-g-PMMA or PS-g-PMMA, respectively.

More recently, an in situ approach has been used to synthesize comb and grafted copolymers and does not require the isolation of the macromonomer as a purification step. The synthesis of the macromonomer involves the slow addition of living polymer to 4-(chlorodimethylsilyl)styrene (CDMSS) (Scheme 11) [109, 110]. This is made possible because of the selectivity of the substitution reaction between the organolithium and the silyl chloride rather than with the styrenic double bond. End capping the living polymer with a few butadiene units before introduction of CDMSS provides greater control as a result of the selectivity for the Si-Cl over the double bond being PBdLi > PILi > PSLi [108]. This method also allows for the incorporation of multifunctional branch points via the copolymerization of “double-tailed” and “triple-tailed” macromonomers, as shown by Hadjichristidis and coworkers [111, 112].

Scheme 11
scheme 11

In situ preparation and copolymerization of macromonomer [110]

Full size image

The use of this in situ macromonomer approach has been extended by Hadjichristidis and coworkers to the synthesis of a host of complex architecture grafted polymer and copolymers (Scheme 12), including comb, star-comb, comb-on-comb, and double-graft architectures (diblock copolymers where each block is a graft copolymer) [113–115].

Scheme 12
scheme 12

Complex comblike block copolymer architectures [115]

Full size image

Foster and coworkers have very recently introduced a new method combining the macromonomer approach and click chemistry for the synthesis of functionalized comb polystyrenes (Scheme 13) [116]. ω-(p-Vinylbenzyl)PS macromere was synthesized by anionic polymerization of styrene, followed by termination with 4-vinylbenzyl chloride. An unsaturated anionic initiator, 4-pentenyllithium, was used to make α-4-pentenyl-ω-(p-vinylbenzyl)PS macromonomer. These two macromonomers were then anionically copolymerized, with the titration of excess sec-BuLi initiator to remove impurities just prior to polymerization, to afford a polymer backbone where the pendant vinyl end groups could be functionalized in a variety of ways using “click chemistry.”

Scheme 13
scheme 13

Mechanism for the synthesis of functionalized comb polystyrenes: (a) macromonomer A, (b) macromonomer B, and (c) synthesis of comb-vinyl copolymer by thiol-ene “click” chemistry [116]

Full size image

3 More Advanced Methods to Achieve Graft Copolymers with Superior Control of Macromolecular Architecture

The methods previously discussed can, under appropriate conditions, allow for control of side chain length and backbone length. However, they provide only statistical control over the number of branch points per molecule and the spacing of the branch points. Thus, much research has been conducted over the past two decades in order to better control these parameters.

Progress in living anionic polymer synthetic techniques has prompted the creation of many novel branched architectures including bottlebrush, π-shaped, H-shaped, super-H-shaped, and pom-pom-shaped polymers and structures incorporating dendritic motifs [20, 117–125]. The synthesis of graft copolymer systems containing regular branch point spacing has also been achieved primarily through the use of chlorosilane-coupling chemistry, originally demonstrated in the synthesis of a simple graft PI-g-PS in 1990 [126]. This material was composed of a polyisoprene (PI) backbone with a single centrally located polystyrene (PS) side chain; the simple graft is equivalent to an A2B miktoarm star. This chemistry has evolved to allow the synthesis of multigraft copolymers having both regular branch point spacing and tunable branch point functionality including tri-, tetra-, and hexafunctional branch points, termed regular comb, centipede, and barbwire architectures, respectively (Scheme 14), as discussed in the next section [127].

Scheme 14
scheme 14

Regular multigraft poly(isoprene-g-styrene) architectures: (a) A(BA) N “comb,” (b) A(B2A) N “centipede,” and (c) A(B4A) N “barbwire” [127]

Full size image

In each case, high molecular weights and a large number of branch points (>10) have been obtained. PI-g-PS materials have been extensively studied and have shown noteworthy superelastic characteristics of extremely high elongation at break and very low hysteresis, with both tensile strength and strain at break increasing with the number of branch points and branch point functionality [128]. More recently, a new class of regularly branched and low polydispersity comb and graft copolymers has been demonstrated and termed “exact” graft copolymers [129]. This novel technique introduces new and unprecedented control of branch point placement, but for practical reasons may be limited to low numbers of branch points (≤5) and moderate molecular weights [9, 130, 131].

3.1 The Chlorosilane Macromonomer Polycondensation Approach

An alternative macromonomer strategy based on step-growth polymerization has been used to make multigraft copolymers with regularly spaced tri-, tetra-, and hexafunctional branch points. As noted above, these materials have been termed “regular combs,” “centipedes,” and “barbwire” polymers, respectively, and have been studied in detail for graft copolymers containing styrene and isoprene [127, 132]. The construction of each multibranched architecture incorporates the same general methodology of combining living anionic and condensation polymerizations, with appropriate choice of the chlorosilane linking agent. In the case of the tetrafunctional branch points (centipedes), living PSLi is slowly added to SiCl4 (“vacuum titration”) to obtain a coupled PS product with two terminal Si-Cl bonds in the middle of the chain. This is then reacted with a small excess of difunctional PI, initiated with a difunctional lithium initiator (DLI) derived from sec-BuLi and (1,3-phenylene)bis(3-methyl-1-phenylpentylidene)dilithium, giving well-defined graft copolymers with a PI backbone and PS branches (Scheme 15). Notice that the last step in the synthesis is a polycondensation reaction yielding a PDI of 2 and allowing the number of branch points to be controlled through stoichiometry.

Scheme 15
scheme 15

Synthesis of PI-g-PS centipedes [132]

Full size image

Correspondingly, tri- or hexa-chlorosilanes are used for the synthesis of comb and barbwire architectures. The hexafunctional multigraft (barbwire) synthesis also requires the strategy of titrating in living PS arms in order to have stoichiometric equivalents of silane and titrant to avoid the substitution of the third Cl at either end of the hexafunctional chlorosilane. In contrast, the synthesis of trifunctional multigrafts requires no titration and is achieved by reacting living PS with an excess of methyltrichlorosilane, followed by removal of excess chlorosilane and addition of living PI having anions at both ends. This method, although painstaking due to the titration step and removal of the excess MeSiCl3 on the high-vacuum line, yields regular branch point spacing, controlled branch point functionality, and is capable of extremely high molecular weights for both homopolymer and copolymer versions [127, 132]. Additionally, narrow molecular weight distribution fractions having high levels of architectural homogeneity were obtained via solvent/nonsolvent fractionation [127, 132].

Plamper et al. recently reported alternate synthetic strategies leading to similar macromolecular architectures, which they called the “pearl necklace architecture” or “threaded star-shaped copolymers” [133]. A polycondensation reaction of PEO macromonomers with partially protected dipentaerythritol leads to a multiblock of PEO. Then the protected hydroxyl groups of dipentaerythritol were removed, allowing attachment of poly(dimethylaminoethyl methacrylate) chains.

3.2 Exact Graft Copolymers

The synthesis of “exact” graft copolymers has been achieved utilizing the macromonomer approach by incorporating moieties based on 1,1-diphenylethylene (DPE). Since DPE shows no self-addition behavior, the dependence on reactivity ratios of the macromonomer and comonomer can be avoided. This technique was first used to synthesize miktoarm stars [134, 135], but its use has now expanded to more complex branched architectures.

Comb, two and three branch symmetric, and asymmetric structures were first shown by Hadjichristidis and coworkers, incorporating 1,4-bis(phenylethenyl)benzene as a macromonomer [136]. This chemistry evolved to produce more elaborate branched structures through the use of 4-(dichloromethylsilyl)diphenylethylene (DCMSPDE). Exact graft copolymer synthesis is highlighted by incorporating two chlorosilane linking molecules, promoting an exchange reaction with living anionic chain ends on a non-homopolymerizable double bond of DPE (Scheme 16). Again, since the substitution reaction is much faster, this allows for complete addition of the living polymer chains without initiation and propagation of the sterically hindered vinyl group. Stoichiometry is crucial in order to obtain complete initiation of the DCMSPDE double bond, without leading to linear side products because of excess sec-BuLi [109, 129, 137].

Scheme 16
scheme 16

Synthesis of exact graft copolymer composed of polystyrene and polyisoprene [130]

Full size image

A second exact grafting strategy demonstrated by Hirao and coworkers involves three reaction steps using a double-DPE macromolecule: (1) a transformation reaction of the α-terminal tert-butyldimethylsilyloxypropyl (SiOP) group into bromopropyl function via deprotection of the SiOP group followed by bromination, (2) a linking reaction of α-SiOP-ω-DPE-functionalized living PS with α-terminal bromopropyl-functionalized PS to prepare an α-SiOP-in-chain-DPE-functionalized PS backbone chain with the introduction of a DPE moiety between the two PS chains, and (3) an addition reaction of PSLi with the DPE moiety to introduce a PS graft chain (Scheme 17) [130].

Scheme 17
scheme 17

Successive synthesis of exact graft PS with three, four, and five PS branches [130]

Full size image

This general strategy was also used in the synthesis of exact graft copolymers having a PMMA backbone and precisely located PS branches (Scheme 18) [131]. This approach was extended to other polymer combinations including P2VP-g-PS, PtBMA-g-PS, and poly(ferrocenylmethyl methacrylate)-g-PS [138]. In these cases, the maximum number of branches attached was five, although in principle more can be achieved if adequate care is taken.

Scheme 18
scheme 18

Synthesis of a series of exact graft copolymers containing PMMA and PS [131]

Full size image

Most recently, Hirao and coworkers have reported the synthesis of exact (co)polymers having six PS, PI, or PMMA grafts attached to a PS backbone [139]. The PS backbone, precisely controlled in terms of DPE placement, number of DPE groups, and molecular weight, was prepared by combining the iterative synthetic sequence with the subsequent coupling reaction and then reacted with the living anions (Scheme 19). In addition, a new double-tailed exact graft PS having 12 PS side chains (two per branch point) was synthesized by repeating the addition reaction of living PS to the backbone PS [139].

Scheme 19
scheme 19

Iterative method allowing exact graft copolymer synthesis up to six branch points [139]

Full size image

4 Characterization

4.1 Molecular Weight Determination of Graft Copolymers

It was previously noted that the rigorous molecular characterization of comb, grafts, and other branched polymers and copolymers can be as arduous as their synthesis. When possible, it is highly desirable to obtain molecular weight (MW) and PDI data on the side chains and backbone prior to linking. Both the “grafting to” and “grafting through” methods allow the backbone and side chains to be characterized independently, as linear constituents. Characterization of the final products (branched polymers and copolymers) is more complicated due to the increase in heterogeneity and their more complex architectures. A prime example of instrument limitations is demonstrated by size exclusion chromatography (SEC). Characterization via SEC is the most commonly utilized approach for MW and PDI determination, but this method is incapable of separating polymers with nearly identical hydrodynamic volumes [140]. Such a restriction is particularly crippling in characterization of model complex branched polymers and copolymers where components of different branching levels, molecular weights, and compositions may all co-elute. Furthermore, SEC suffers from large band broadening effects, which make it even more difficult to resolve the various species present in complex branched polymers and copolymers.

While synthetic polymer chemists continue to develop better synthetic methodologies that reduce structural heterogeneity, finding alternatives to SEC for the rigorous molecular weight characterization of highly branched and other nonlinear structures is a necessity. Where structural perfection is currently not possible, characterization of imperfect structures can be just as beneficial [141]. In the last 15 years, this has been made feasible with the introduction of temperature gradient interaction chromatography (TGIC) and first demonstrated by Chang et al. [142]. TGIC relies on the enthalpic interactions between the solute molecules and the stationary phase; these interactions can allow separation based on molecular weight and not hydrodynamic volume. Strict control of the solvent and temperature leads to superior resolution of branched polymers when compared to SEC with far reduced band broadening [141, 143, 144].

A recent review by Hutchings has extensively discussed the advantages and limitations of TGIC as it pertains to numerous complex branched architectures including comb, H-shaped, and particular emphasis on dendritically branched polymers [141]. In all cases, the use of TGIC was able to show that impurities were present after fractionation, whereas traditional SEC analysis showed only a narrow, single peak. More specifically, in the case of H-shaped polymers, it has been shown that TGIC is capable of removing uncertainty and allowing for validation of current rheological models for branched polymers [145–147]. TGIC will continue its advancement and become a more universal characterization technique as a result of its superior resolution and ability to identify low levels of heterogeneity.

4.2 Morphology and Mechanical Properties of Graft Copolymers

Block copolymers undergo phase separation and self-organization on different length scales, ranging from nanometers to hundreds of nanometers as a result of molecular weight, block composition, the solvent for film casting, annealing, etc. [6]. For linear diblock and triblock copolymers, the morphology (spheres, cylinders, bicontinuous gyroid, and lamellae) is directly linked to component volume fractions. However, until about 20 years ago, very little was known about how long-chain branching impacts morphology and mechanical properties of block copolymers.

Milner developed a self-consistent mean field model for effects of architectural and conformational asymmetry on “opposing polymer brushes,” which approximate miktoarm star copolymers [148]. Milner predicted that changing architecture from an AB diblock to A2B, A3B, and A4B miktoarm stars, while keeping composition constant, would systematically alter the morphology of these materials. Thus, Milner predicted that macromolecular architecture could be used to decouple morphology of copolymers from the hitherto observed strict dependence on composition [148]. This prediction was subsequently verified [149–151].

In order to apply Milner’s theory to more complex multigraft copolymers, the constituting block copolymer hypothesis (Scheme 20) is employed. This concept is based on the idea that the overall phase behavior of a grafted copolymer is governed by the behavior of smaller copolymer units associated at each junction point [123, 152–154]. Existing theories of miktoarm star and asymmetric linear diblocks can thus be used to predict the behavior of the overall graft copolymers based upon the local resemblance of multigraft copolymers to miktoarm star copolymers [148, 155–157].

Scheme 20
scheme 20

Constituting miktoarm star concept [159]

Full size image

Multigraft systems containing increased branch points per molecule have been shown to suppress long-range ordering (Fig. 1) [128, 159]. Transmission electron microscopy (TEM) images can only provide hints to the microphase-separated domains for these materials [127]. Small-angle X-ray scattering (SAXS) data supply confirmation of morphology and allow for the calculation of periodicity of microphase separation from peak position (q) by application of the Bragg equation [158]. Thus, when TEM observations are inconclusive because the material is disrupted by branching or by alternative chemical/architectural means, SAXS can yield well-reasoned morphological interpretations.

Fig. 1
figure 1

TEM images of hexafunctional multigraft copolymers containing 21 % PS with (a) 5.3, (b) 3.6, and (c) 2.7 junctions per molecule. Additionally, SAXS confirmed (a) microphase-separated and (b, c) lamellae morphologies [128] (Reprinted with permission from Macromolecules (2006) American Chemical Society)

Full size image

Uniformity of branch placement has been demonstrated to control the morphology of the molecule at large through the preferred behavior at the smaller subunits of the junction points. Beyer et al. showed that the domain shape of PI-g-PS with tri- and tetrafunctional branch points can be correctly predicted, with the lamellar morphology being the only to give long-range ordering [159]. A second finding demonstrated in this work was that branch point placement was more significant than the polydispersities of each sample [159]. Supporting these results, graft copolymers with random tri- and tetrafunctional branch junction placement displayed characteristic microphase separation but no long-range ordering as a consequence of different morphologies being preferred throughout the polymer chain [154].

Multigraft copolymers also exhibit unique mechanical properties emphasizing the influence of molecular architecture. Again, well-controlled PI-g-PS composed of regularly spaced grafted structure varying in the number of branch junctions per molecule and branch point functionality (synthesized anionically under high-vacuum conditions) was investigated for their stress-strain behavior (Fig. 2) [128, 160]. Analysis showed that the strain at break far exceeded that of their linear counterparts (Styroflex and Kraton) [161]. Both strain at break and tensile strength showed a linear dependence on the number of branch points. Hysteresis upon stretching was extremely small, and the modulus could be tuned at a given composition by varying branch point functionality [162]. PI-g-PS structures composed of random branch points yielded similar results, demonstrating the superior properties of multigraft copolymers as a result of their molecular architecture [128, 162, 163]. As the synthesis and characterization of graft copolymer and other complex architectures continue to progress, these materials can continue to be tailored to optimize physical properties and exploit control of macromolecular architecture for new and existing applications.

Fig. 2
figure 2

(a) Stress-strain behavior for Styroflex, Kraton, and two tetrafunctional multigraft copolymers (PI-g-PS). (b) Influence of junction point functionality on strain at break [128] (Reprinted with permission from Macromolecules (2006) American Chemical Society)

Full size image

5 Conclusions and Future Prospects

Living anionic polymerization is an extremely valuable and versatile technique for the synthesis of well-defined comb polymers and graft copolymers, both alone and in combination with other living polymerization techniques. Over the past two decades, enormous advances in polymerization methodologies have led to comb and graft materials having a broad range of chemical compositions and applications ranging from elastomers and tough plastics to biomaterials and pharmaceutical applications. New synthetic strategies allow control, not only of backbone and side chain MW and PDI but also of the exact number of branch points, their placement along the backbone, and the number of branches per branch point.

TGIC is emerging as a key analytical technique for the rigorous molecular characterization of complex branched polymers and copolymers. Superior mechanical properties have been demonstrated for multigraft copolymer-based thermoplastic elastomers.

In the future and following a general trend in block copolymer research, we anticipate additional work in the synthesis of more complex multicomponent graft copolymers, incorporating three or more monomers. We also anticipate the use of self-assembly in creating reversible multigraft copolymers that respond to external stimuli. Polypeptide components may play a key role here.