Nitroxide mediated controlled synthesis of glycidyl methacrylate-rich copolymers enabled by SG1-based alkoxyamines bearing succinimidyl ester groups

Abstract

The controlled nitroxide-mediated copolymerization of glycidyl methacrylate (GMA) and styrene (S) with varying GMA molar feed fractions (f_GMA,0 = 0.12–0.94) was accomplished by using a SG1-based alkoxyamine initiator bearing a N-succinimidyl ester group (NHS-BlocBuilder) in 50 wt% 1,4-dioxane solution at 90 °C. Copolymerizations indicated linear evolution of number average molecular weight M_n with respect to conversion up to approximately 50% and narrow molecular weight distributions with M_w/M_n = 1.22–1.44 and GMA incorporation into copolymer (F_GMA) as high as 0.92. No additional SG1 free nitroxide was required to control polymerizations, even at high f_GMA,0. Chain extensions of poly(GMA-ran-S) macroinitiators with S at 110 °C yielded a high fraction of block copolymer in most cases (except at the highest F_GMA), as clear, monomodal shifts in M_n using gel permeation chromatography (GPC) were observed, thereby suggesting the poly(GMA-ran-S) macroinitiators were substantially “living”.

---

Introduction

Controlled radical polymerization (CRP) has been a major achievement in polymer chemistry over the past 15 years because of its potential to engineer materials with narrow molecular weight distribution and controlled microstructure, without the stringent requirements to obtain such polymers by classical “living” polymerizations. All CRP methods are based on the reversible activation–deactivation equilibrium between active and dormant species, which controls the concentration of active radicals such that irreversible termination reactions are suppressed during a significant portion of the polymerization.^1–3

Atom transfer radical polymerization (ATRP),^4–6nitroxide-mediated polymerization (NMP),^7,8 and reversible addition–fragmentation chain transfer (RAFT)^9,10 are common CRP methods and have been used to make next-generation biomaterials,^11–13 organic electronics,¹⁴ and nanoporous materials.^15–20 Initially, NMP was only applicable to styrenic monomers using 2,2,6,6-tetra-methyl-1-piperidinyloxy (TEMPO) mediator.^21,22 Second-generation nitroxides like the so-called SG1 (N-(2-methylpropyl)-N-(1-diethylphosphono-2,2-dimethylpropyl)-N-oxyl) and the related BlocBuilder that is sold commercially by Arkema (N-(2-methylpropyl)-N-(1-diethylphosphono-2,2-dimethylpropyl)-O-(2-carboxylprop-2-yl) hydroxylamine)²³ (Scheme 1) and TIPNO²⁴ (2,2,5-trimethyl-4-phenyl-3-azahexane 3-nitroxide), however, enabled controlled polymerization of acrylates, acrylamides, and to some extent, methacrylates.^25–27

Scheme 1 Illustration of the nitroxide initiators: (a) BlocBuilder is based on the SG1 family of nitroxide; (b) BlocBuilder is reacted with N-hydroxysuccinimide to form the succinimidyl ester terminated BlocBuilder that is used in this study.

The key problem when using NMP to control the polymerization of methacrylic monomers such as methyl methacrylate (MMA) is due to the cross-disproportionation and large activation–deactivation equilibrium constant (K) associated with methacrylates.^28,29 The polymerization of MMA with TEMPO failed because of a cross-disproportion side reaction due to β-hydrogen transfer from propagating radicals to initiator.³⁰ However, with a nitroxide such as SG1, cross-disproportionation side reactions have not been observed at low levels³¹ but excessive SG1 led to cross-disproportionation.³² Even with low levels of SG1, MMA polymerization was still uncontrolled due to the large K.³¹ The problem with the large K can be mitigated by the co-monomer approach for BlocBuilder/SG1 type initiators suggested by Charleux and co-workers.²⁷ This method relies on copolymerizing the methacrylate with a small fraction of co-monomer with a much lower K, such as styrene. This effectively reduces the average K to such a level that controlled polymerizations of essentially pure methacrylic polymers can be achieved by NMP.^27,29 It should be noted that other authors have recently developed nitroxides that can homopolymerize methacrylates although chain extension to form block copolymers in some cases was problematic.^33–38

Recently, Vinas et al. have developed an initiator based on the commercially available BlocBuilder alkoxyamine bearing a succinimidyl ester moiety (NHS-BlocBuilder) (Scheme 1).³⁹ They have shown the capability of this new initiator by synthesizing well-defined α-NHS functional poly(styrene) and poly(n-butyl acrylate) which, after suitable transformation of the NHS terminal group, could be used as a macroinitiator for the ring-opening polymerization of lactide to form diblock copolymers. The versatility of the NHS moiety has been demonstrated by its application towards attaching amine-functional molecules such as polypeptides or inorganic fillers to polymers.^40,41 The NHS-BlocBuilder presents interesting possibilities for controlling NMP as it was found to have a much higher dissociation constant compared to BlocBuilder (∼15 times), so that it effectively provides for sufficient excess SG1 to control styrene and n-butyl acrylate polymerizations from the onset. This feature has not been reported for methacrylate polymerizations with BlocBuilder, which typically requires ∼10 mol% excess SG1 relative to BlocBuilder for a controlled polymerization.²⁹

Glycidyl methacrylate (GMA) containing copolymers are of interest for their epoxy group⁴² that can be used to cross-link domains, for example, in ordered block copolymers for nanoporous membranes.^16,17,19GMA has also been used for homogeneous and heterogeneous polymer networks⁴³ and for coatings, matrix resins and adhesives.^44,45CRP of GMA by ATRP^43,46 and RAFT⁴⁷ has been investigated but literature is limited regarding GMA by NMP.^24,48 Benoit et al. copolymerized the related monomer glycidyl acrylate (GA) with styrene using a TIPNO-based initiator for GA feed concentrations <20 mol%, resulting in resins with number average molecular weights M_n = 20–23 kg mol⁻¹ and narrow molecular weight distributions (M_w/M_n = 1.16–1.18).²⁴ However, higher feed concentrations ∼50 mol% GA resulted in higher M_w/M_n = 1.52. Grubbs et al. copolymerized one composition of GMA with methyl acrylate (20 mol% GMA in the feed) using TIPNO, resulting in a copolymer with M_n = 6.6 kg mol⁻¹ and M_w/M_n = 1.25.⁴⁸ Our group earlier was able to successfully incorporate GMA at one particular composition into a block copolymer by chain extension from a poly(methacrylate-ran-styrene) macroinitiator (poly(MMA-ran-S)) using BlocBuilder.⁴⁹

Applying BlocBuilder to control GMA-containing feeds is problematic as the epoxy functionality in the monomer could react with the carboxylic acid attached to BlocBuilder, leading to branched, insoluble products. Indeed, the epoxy/carboxylic acid reaction has long been widely used in curing systems⁵⁰ and in reactive compatibilization.^51,52 Our earlier work did not indicate any sort of premature cross-linking, likely due to the dilution of the acid end group from the macroinitiator. Higher GMA concentrations in the second batch however resulted in insoluble products.⁵³ Therefore, protection of the acid group on BlocBuilder was deemed necessary and we chose to use NHS-BlocBuilder, as it should not be reactive with the epoxy group. Thus, this study proposes to examine the ability of NHS-BlocBuilder to control a wide range of GMA/styrene feed compositions. Further, the resulting copolymers will be tested for their ability to reinitiate a second batch of monomer to show if block copolymers can be formed from such macroinitiators. Finally, all of the copolymerizations were done without any added SG1 free nitroxide, which was thought to be necessary to effectively control methacrylate-rich copolymerizations by BlocBuilder/NMP. This latter feature will greatly simplify further work with methacrylate NMP processes as it potentially eliminates the need for extra nitroxide in the formulation.

Experimental

Materials

Calcium hydride (90–95% reagent) and basic alumina (Brockmann, Type 1, 150 mesh) were obtained from Sigma-Aldrich and used as received. Tetrahydrofuran (THF, 99.9% HPLC grade), dioxane (99%), methanol (99%), p-xylene (99.9%), heptane (98%) and hexane (95%) were received from Fisher Scientific and used as received. N-(2-Methylpropyl)-N-(1-diethylphosphono-2,2-dimethylpropyl)-O-(2-carboxylprop-2-yl) hydroxylamine (99%, BlocBuilder™) was received from Arkema. N-Hydroxysuccinimide (98%) and N,N′-dicyclohexylcarbodimide (DCC, 99%) were received from Sigma-Aldrich and used in conjunction with BlocBuilder to synthesize the succinimidyl ester terminated alkoxyamine BlocBuilder (NHS-BlocBuilder) using the same procedure as Vinas et al.³⁹Styrene (S, 99%) and glycidyl methacrylate (GMA, 97%) were both obtained from Sigma-Aldrich and were purified to remove the inhibitor by passing through a column of basic alumina mixed with 5 weight% calcium hydride and then stored in a sealed flask under a head of nitrogen in a refrigerator until needed. The deuterated chloroform (CDCl₃, 99%) used as a solvent for proton nuclear magnetic resonance (¹H NMR) was obtained from Cambridge Isotopes Laboratory.

Testing for cross-linking between glycidyl methacrylate and BlocBuilder

An equimolar mixture of GMA was reacted with the BlocBuilder in xylene solution to test for the cross-linking of GMA with BlocBuilder. Another experiment using MMA with BlocBuilder was also done as a control experiment. The reactions were done in a 50 mL three-neck round bottom glass flask equipped with a reflux condenser, thermal well and a magnetic stir bar. The flask was placed inside a heating mantle and the whole set-up was installed on the top of a magnetic stirrer. Table 1 shows the formulations used. An example is given to illustrate the reaction of GMA with BlocBuilder. To the reactor were added BlocBuilder (0.19 g, 0.50 mmol), GMA (0.064 g, 0.45 mmol) and xylene (6.90 g, 65 mmol). The reactor was then sealed and purged with nitrogen for 30 minutes at ambient temperature. Then, the heating was increased to 90 °C at a rate of about 10 °C min⁻¹ while continuing the purge. Once the set point temperature was reached, heating continued for 1 h. Then, heating was stopped and once cooled to <35 °C, the contents were precipitated into heptane and dried under vacuum at low temperature overnight. The sample was then placed in an NMR tube, filled with CDCl₃ and then analysed by ¹H NMR.

Table 1 Compositions for methacrylate/BlocBuilder equimolar reactions at 90 °C and 1 hour

Expt. ID	[XMA]^a/M	[BlocBuilder]/M	[Xylene]/M
a XMA refers to the methacrylate where X = glycidyl (G) or methyl (M).
GMA/BlocBuilder	0.69	0.75	8.0
MMA/BlocBuilder	0.69	0.70	8.0

---

Synthesis of glycidyl methacrylate/styrene random copolymers

The syntheses were conducted in a 100 mL three-neck round bottom glass flask equipped with a reflux condenser, thermal well and a magnetic stir bar. The flask was placed inside a heating mantle and the whole set-up was installed on the top of a magnetic stirrer. Table 2 lists the formulations that were studied for the GMA/S copolymerization. As an example, for the experiment GMA/S-50, NHS-BlocBuilder (0.20 g, 0.42 mmol), purified S (4.38 g, 42 mmol), purified GMA (6.07 g, 43 mmol) and 1,4-dioxane (10.37 g, 118 mmol) were added to the reactor and then the reactor was sealed with rubber septa. The thermocouple was placed inside the thermal well connected to one of the necks and then it was connected to the temperature controller. A mixture of glycol/water (90/10 vol%) at a temperature of 0 °C was circulated (Fisher Scientific Isotemp 3006D refrigerating circulator) through the condenser connected to one of the necks of flask to prevent any evaporation of monomers and/or solvent. A purge of ultra pure nitrogen was then introduced to the reactor for 30 minutes at ambient conditions in order to deoxygenate the reaction environment prior to polymerization. The purge was vented through the reflux condenser.

Table 2 Formulations for glycidyl methacrylate/styrene random copolymerizations initiated by N-hydroxysuccinimidyl BlocBuilder at 90 °C in 1,4-dioxane solution

Expt. ID	f GMA,0	[Initiator]0/M	[GMA]0/M	[ST]0/M	[Dioxane]0/M
GMA/S-10	0.12	0.019	0.5	4.0	5.6
GMA/S-20	0.20	0.019	0.9	3.5	5.6
GMA/S-30	0.30	0.018	1.3	3.0	5.6
GMA/S-40	0.43	0.018	1.7	2.3	5.9
GMA/S-50	0.50	0.020	2.1	2.0	5.7
GMA/S-60	0.59	0.019	2.4	1.7	5.7
GMA/S-70	0.71	0.020	2.8	1.1	5.7
GMA/S-85	0.84	0.019	3.4	0.6	5.5
GMA/S-90	0.90	0.019	3.4	0.4	5.9
GMA/S-95	0.94	0.019	3.5	0.2	5.8

---

After purging at room temperature for 30 minutes, the heating of the reactor contents was commenced at a rate of 10 °C min⁻¹ to 90 °C while maintaining the nitrogen purge. The time at which the reactor temperature reached 90 °C was taken as the start of the reaction. Samples were taken from the reactor periodically by a syringe until the samples became too viscous to withdraw. Reactions were stopped by removing the reactor from the heating mantle when the mixtures became noticeably viscous. Reaction time varied greatly depending on the initial feed composition (420 min for f_GMA,0 = 0.12 and 20 min for f_GMA,0 = 0.94). For each sample withdrawn during the polymerization, the polymer was precipitated by excess hexane. After filtration and recovery, the precipitated polymer was dried at 70 °C under vacuum in the oven for 24 hours in order to remove any solvent and unreacted monomers. Samples were re-dissolved again in THF and re-precipitated in excess hexane twice more to remove any further remaining solvent and unreacted monomers. The samples were then dried again in the vacuum oven. The final yield for the experiment GMA/S-50 was 2.93 g (28%). The molar composition of GMA in the copolymer was F_GMA = 0.32, according to ¹H NMR in CDCl₃. The number average molecular weight, M_n, was 8.6 kg mol⁻¹ and polydispersity index M_w/M_n = 1.32 according to gel permeation chromatography (GPC) relative to narrow distribution, linear poly(styrene) standards after suitable composition corrections for the copolymer. THF was the eluent at 40 °C (see Characterization section for further details).

Chain extensions with styrene

To examine the “livingness” of the copolymers, chain extensions with styrene were performed using the copolymers synthesized in the previous section as macroinitiators. For all chain extensions, the theoretical molecular weight of the extended polymer was set to ∼100 kg mol⁻¹ in order to observe clear molecular weight shifts from the macroinitiator to the chain-extended species. The formulations for the chain extension experiments are listed in Table 3.

Table 3 Formulations of styrene chain extensions from various glycidyl methacrylate/styrene random copolymer macroinitiators in 50 wt% 1,4-dioxane solution at 110 °C

Expt. ID	[I]0^a/mmol	M n,GMA/S/kg mol^−1	n S,0 /mmol	n dioxane /mmol
a [I]0 refers to the initial moles of macroinitiator added.
GMA/S-10-S	0.056	8.9	59	71
GMA/S-50-S	0.060	8.6	50	66
GMA/S-95-S	0.051	0.0	50	63

---

These experiments were conducted in the same reactor apparatus as used for the macroinitiators. An example is given to illustrate the chain extension of GMA/S macroinitiator with a fresh batch of S. For the experiment GMA/S-50-S, to the reactor were added GMA/S-50 macroinitiator (0.50 g, M_n = 8.6 kg mol⁻¹, M_w/M_n = 1.30, F_GMA = 0.32), purified S (5.15 g, 50 mmol) and 1,4-dioxane (5.77 g, 66 mmol). Samples were periodically removed by syringe and were precipitated using excess hexane. Polymerization was run until the reaction mixture became noticeably more viscous (for the aforementioned experiment, GMA/S-50-S, this occurred at about 240 min). After precipitation into hexane and recovery, the crude product was dried under vacuum at 70 °C for 24 hours in order to obtain the GMA/S-50-S block copolymer with a final yield of 1.86 g (conversion of S block = 0.35, M_n = 47.6 kg mol⁻¹, M_w/M_n = 1.44, F_GMA = 0.04).

For the chain extension of S from GMA/S-95, noticeable “dead” macroinitiator was observed by GPC and fractionation was done to separate the macroinitiator from the chain-extended product. The crude copolymer GMA/S-95-S was dissolved in dioxane (1 g per 15 mL dioxane). A 50/50 v/v mixture of methanol/de-ionized water was added slowly until precipitation was noticeable. The crude polymer was retrieved and fractionated again, dried and characterized for composition by ¹H NMR and by GPC for molecular weight analysis.

Characterization

Number average molecular weight (M_n), weight average molecular weight (M_w) and polydispersity index (M_w/M_n) were determined by gel permeation chromatography (GPC) which was calibrated relative to linear poly(styrene) standards with THF as the eluent at 40 °C. A Waters Breeze GPC system was used at a mobile phase flow rate of 0.3 mL min⁻¹ equipped with three Styragel HR columns (HR1 with a molecular weight measurement range of 10² to 5 × 10³ g mol⁻¹, HR2 with a molecular weight measurement range of 5 × 10² to 2 × 10⁴ g mol⁻¹and HR4 with a molecular weight measurement range of 5 × 10³ to 6 × 10⁵ g mol⁻¹) and a guard column. The GPC was equipped with a Waters 2487 UV detector set at a wavelength of 255 nm to detect aromatic groups in poly(styrene) containing copolymers and an RI 2410 differential refractive index (RI) detector. For these experiments, the RI detector was used. The copolymer molecular weights were corrected using the Mark–Houwink relationship [η] = KM^α, based on the following Mark–Houwink coefficients:⁵⁴α_PS = 0.716, α_PGMA = 0.537, K_PS = 1.14 × 10⁻⁴ dL g⁻¹ and K_PGMA = 2.78 × 10⁻⁴ dL g⁻¹. The molecular weights were corrected by compositionally averaging the Mark–Houwink coefficients for the copolymers and then correcting appropriately against the pure poly(styrene) standards.⁵⁵

¹H NMR was used to determine the copolymer composition. A 300 MHz Varian Gemini 2000 spectrometer was used for the ¹H NMR measurements. Samples were placed in 5 mm up NMR tubes using CDCl₃ as a solvent. After injecting and shimming, the samples were scanned for 32 times in the NMR apparatus. The GMA/S copolymer composition was found by using the resonances due to the aromatic protons of S at δ = 6.8–7.3 ppm and one of the methylene protons of GMA at δ = 4.3–4.5 ppm.⁵⁴ Other suitable resonances from the GMA protons could be used as well (e.g. δ = 3.3 ppm, 2.9 ppm and 2.7 ppm). The mole fraction of GMA in the copolymer was calculated according to F_GMA = b/(b + a/5) where a and b are the resonance peak areas due to the aromatic protons from the styrene units and the methylenic protons from the GMA units, respectively. The overall monomer conversion was determined by gravimetry. It should be noted that using this technique may contain some error as precipitation and recovery of polymer may eliminate some oligomers, which could be soluble in the precipitating solution.

Result and discussion

Choice of NHS-BlocBuilder as controlling nitroxide

Initial experiments using the carboxyl terminated BlocBuilder and free nitroxide to control GMA/S copolymerizations resulted in insoluble products, presumably due to the carboxyl group reacting with the epoxy group in GMA. Thus, we attempted to use ¹H NMR to identify the products formed by reacting equimolar GMA/BlocBuilder mixtures and as a control, an equimolar mixture of non-functional MMA and BlocBuilder was also reacted (see Scheme 2). If the epoxy/acid reaction occurred, then the vinylic protons of GMA should be noticed at δ ≈ 5.5 to 6 ppm. If the esterification reaction was occurring, then the vinylic protons would not show in the spectrum. Both reactions were done for the same duration of time (1 h) and temperature (90 °C).

Scheme 2 Illustration of side reaction possible between the epoxy functional glycidyl methacrylate (GMA) monomer and the carboxylic acid functional BlocBuilder initiator.

The control experiments indicated no vinylic protons were present after 1 h, suggesting adequate time was allotted for the addition across the double bond to form the MMA/BlocBuilder adduct. In contrast, vinylic protons were observed for the GMA/BlocBuilder experiment (Fig. 1). If only addition was occurring, it would be expected that the vinyl groups would have been completely consumed, since the propagation rate constant for GMA is higher compared to that of MMA.⁵⁶ Thus, the epoxy/acid reaction was competing with the desired reaction.

Fig. 1 ¹H NMR spectrum of the glycidyl methacrylate/BlocBuilder product after 1 h at 90 °C. Note that the vinylic protons are still present while there are strong resonances at about 4–4.5 ppm that is suggestive of ester formation.

To restrict the interference of the acid with the epoxy functional monomer, protection schemes for BlocBuilder were sought. We chose to protect the acid by reacting BlocBuilder with N-hydroxy succinimide (NHS) to form the succinimidyl ester terminated BlocBuilder (NHS-BlocBuilder). Several reasons make NHS-Builder an attractive choice to control the polymerization. It renders, for example, the acid non-reactive during the polymerization. After the polymerization, the NHS could be reacted with amines, for example, to further functionalize the chains. Lastly, the rapid dissociation of NHS-BlocBuilder compared to BlocBuilder permitted polymerizations to be done without any added free nitroxide at the onset of the polymerization. This was shown for n-butyl acrylate³⁹ and our study attempted to determine if this same protocol was possible for methacrylate-rich feeds.

Random copolymerization of glycidyl methacrylate/styrene: effect of initial feed composition on copolymerization kinetics

As summarized in Table 2, GMA/S copolymerizations were conducted in 1,4-dioxane solution at 90 °C using NHS-BlocBuilder with GMA feed compositions f_GMA,0 = 0.12–0.94. Fig. 2 shows the kinetic plots of ln [(1 − X)⁻¹] (X = conversion) versus time of copolymerization for a range of feed compositions. The slopes of these plots were calculated typically from about four to six sample points taken in the linear region. The slopes provide the apparent rate constant, 〈k_p〉[P˙], where 〈k_p〉 is the propagation rate constant and [P˙] is the concentration of the propagating macro-radicals. As Fig. 2 indicates, the kinetic plots were quite linear with respect to copolymerization time for all feed compositions of GMA in the times studied and showed that 〈k_p〉[P˙] increased with increasing concentrations of GMA in the feed. Fig. 3 summarizes more succinctly the influence of feed composition on 〈k_p〉[P˙]. A slow increase was observed as f_GMA,0 increased and only at very high f_GMA,0 >0.8 was there a significant increase in 〈k_p〉[P˙]. This is in agreement with other studies where a rapid increase in 〈k_p〉[P˙] was observed for MMA-rich feeds at high MMA initial feed concentrations >80 mol% using BlocBuilder.²⁷ The rapid increase is accentuated by the difference in the k_p between GMA and S. GMA has a very high k_p compared to S at 90 °C, with k_p,GMA = 3 × 10³ L mol⁻¹s⁻¹ and k_p,S = 9 × 10² L mol⁻¹s⁻¹.⁵⁶ The k_p of GMA is higher compared to MMA (k_p,MMA ≈ 1.6 × 10³ L mol⁻¹s⁻¹)⁵⁶ and no significant exotherms were observed, suggesting that NHS-BlocBuilder was encouraging as a controller for NMP of methacrylic monomers.

Fig. 2 Representative kinetic plots of ln [(1 − X)⁻¹] (X = conversion) versuspolymerization time for glycidyl methacrylate/styrene (GMA/S) random copolymerizations at various initial GMA feed concentrations (f_GMA,0): f_GMA,0 = 0.12 (open squares, □); f_GMA,0 = 0.20 (filled squares, ■); f_GMA,0 = 0.30 (open circles, ○); f_GMA,0 = 0.43 (filled circles, ●); f_GMA,0 = 0.50 (open triangles, △); f_GMA,0 = 0.59 (filled triangles, ▲); f_GMA,0 = 0.71 (open diamonds, ◇); f_GMA,0 = 0.84 (filled diamonds,◆); f_GMA,0 = 0.90 (plus signs, +); f_GMA,0 = 0.94 (cross signs, ×). All polymerizations were done at 90 °C in 50 wt% dioxane solutions with NHS-BlocBuilder unimolecular initiator.

Fig. 3 Plot of apparent rate constant k_p[P˙] (k_p = average propagation rate constant, [P˙] = concentration of propagating macroradicals) as a function of various initial glycidyl methacrylate (GMA) feed concentrations f_GMA,0 for GMA/styrene (S) copolymerizations initiated by NHS-BlocBuilder at 90 °C in 50 wt% dioxane solutions. The k_p[P˙] was obtained from the slopes of the kinetic plots shown in Fig. 2 in the linear regions. Error bars are derived from the standard errors of the best-fit slope from the kinetic plots. The k_p[P˙] (dashed line) was also predicted using the penultimate model using the average propagation rate constant 〈k_p〉 and Fischer's expression for [P˙]². The derivation of the expression is described in detail in the ESI†.

Effect of initial feed composition on molecular weight distribution and copolymer “livingness”

One of the distinct features of CRP compared to conventional radical polymerization is the linear evolution of molecular weight with respect to conversion, which is superficially similar to that of ionic and other truly “living” polymerizations. Further, the molecular weight distribution should be relatively narrow with polydispersity indices (M_w/M_n) < 1.5 and the chain end should be sufficiently functional so as to be able to initiate a second batch of monomer cleanly.

The number average molecular weight (M_n), weight average molecular weight (M_w) and M_w/M_n of each sample taken during the GMA/S copolymerizations were analyzed by GPC and the relationship of measured M_n with conversion was compared to the theoretical values (M_{n,theoretical}) as shown in Fig. 4. M_ns measured by GPC were based on poly(styrene) standards which were then converted to poly(glycidyl methacrylate)/poly(styrene) copolymerM_ns by the Mark–Houwink equation as described in the Characterization section. The molecular characteristics of the various GMA/styrene random copolymers are tabulated in Table 4. In some cases, higher M_n at low conversion compared to M_{n,theoretical} resulted mainly from some soluble oligomers being washed out during precipitation and decanting of samples. Therefore, the loss of these oligomers resulted in slightly higher M_n at low conversions in some cases. Also, the results shown in Fig. 4 suggest that the experimental data resembled the theoretical line trend better at low conversions (<40%) and deviated at higher conversions. The S-rich samples, especially at low conversion, agreed very well with the theoretical prediction. The GMA-rich samples (f_GMA,0 > 0.7) tend to deviate more from the theoretical line and the deviation became more pronounced at higher conversion for the GMA-rich copolymers.

Table 4 Molecular characteristics of the various poly(glycidyl methacrylate-ran-styrene) copolymers synthesized using NHS-BlocBuilder at 90 °C in 50 wt% 1,4-dioxane solution

Expt. ID	f GMA,0	t polymerization /min	conversion	F GMA	M n/kg mol^−1	M w/Mn
GMA/S-10	0.12	420	0.25	0.03	8.9	1.25
GMA/S-20	0.20	240	0.18	0.06	3.9	1.35
GMA/S-30	0.30	220	0.37	0.11	12.4	1.39
GMA/S-40	0.43	290	0.60	0.17	14.9	1.22
GMA/S-50	0.50	80	0.33	0.32	8.6	1.32
GMA/S-60	0.59	110	0.57	0.49	13.8	1.28
GMA/S-70	0.71	80	0.53	0.53	12.4	1.31
GMA/S-85	0.84	55	0.58	0.75	13.4	1.35
GMA/S-90	0.90	62	0.88	0.90	18.5	1.44
GMA/S-95	0.94	20	0.59	0.92	11.4	1.38

---

Fig. 4 Number average molecular weight M_nversus conversion (X) for glycidyl methacrylate/styrene (GMA/S) random copolymerizations at various initial GMA feed concentrations (f_GMA,0): f_GMA,0 = 0.12 (open squares, □); f_GMA,0 = 0.20 (filled squares, ■); f_GMA,0 = 0.30 (open circles, ○); f_GMA,0 = 0.43 (filled circles, ●); f_GMA,0 = 0.50 (open triangles, △); f_GMA,0 = 0.59 (filled triangles, ▲); f_GMA,0 = 0.71 (open diamonds, ◇); f_GMA,0 = 0.84 (filled diamonds, ◆); f_GMA,0 = 0.90 (plus signs, +); f_GMA,0 = 0.94 (cross signs, ×). All polymerizations were done at 90 °C in 50 wt% dioxane solutions with N-hydroxysuccinimidyl BlocBuilder unimolecular initiator.

For GMA/S copolymerizations, the M_w/M_ns (Fig. 5) were broader for samples taken at low conversion but become progressively narrower as the conversion increased. This happens due to the equilibrium between dormant and propagating species not being established sufficiently for the persistent radical effect to be present. However, as conversion increased, the M_w/M_ns decreased and levelled to about 1.3–1.4. Also, it is notable that the M_w/M_ns generally were higher for the copolymers richer in GMA, indicating that it is more difficult to control the methacrylate, as expected. Still, most of the M_w/M_ns were ∼1.3 to 1.4 and this observation was encouraging for the ability of the GMA/S copolymers to reinitiate a second batch of monomer.

Fig. 5 Polydispersity index M_w/M_nversus conversion (X) for glycidyl methacrylate/styrene (GMA/S) random copolymerizations at various initial GMA feed concentrations (f_GMA,0): f_GMA,0 = 0.12 (open squares, □); f_GMA,0 = 0.20 (filled squares, ■); f_GMA,0 = 0.30 (open circles, ○); f_GMA,0 = 0.43 (filled circles, ●); f_GMA,0 = 0.50 (open triangles, △); f_GMA,0 = 0.59 (filled triangles, ▲); f_GMA,0 = 0.71 (open diamonds, ◇); f_GMA,0 = 0.84 (filled diamonds, ◆); f_GMA,0 = 0.90 (plus signs, +); f_GMA,0 = 0.94 (cross signs, ×). All polymerizations were done at 90 °C in 50 wt% dioxane solutions with N-hydroxysuccinimidyl BlocBuilder unimolecular initiator.

Chain extension of glycidyl methacrylate/styrene random copolymers with styrene

Chain extensions with styrene were performed to examine the ability of the GMA/S macroinitiators synthesized with the NHS-BlocBuilder initiator to cleanly initiate a second batch of monomer. The chain extensions were conducted using copolymers with very different compositions (F_GMA = 0.03, 0.32 and 0.92 denoted by experiments GMA/S-10, GMA/S-50 and GMA/S-95, respectively) as macroinitiators for styrene polymerization at 110 °C. The M_n and M_w/M_n of the samples were analyzed by GPC and compared to those of the macroinitiators. The GPC chromatograms of the various chain extensions are shown in Fig. 6. The changes in M_n, M_w/M_n as well as composition of the block copolymer after the chain extension of macroinitiators with styrene are summarized in Table 5. The M_w/M_n after chain extension increased slightly for the macroinitiators with the lower GMA contents. As the GPC chromatograms in Fig. 6 indicate, clear shifts of M_n were observed for GMA/S-10-S and GMA/S-50-S, which implied nearly simultaneous growth of all chains, with little dead macroinitiator. However, a slight increase in M_w/M_n for these two copolymers was observed which indicates that there was likely a small fraction of dead chains that were present in the macroinitiators (Fig. 6a and b). It is clear that for the chain extension from the macroinitiator with the highest GMA content shown in Fig. 6c, the fraction of dead chains was quite high (∼30% based on fitting the two peaks to Gaussian curves). Thus, GMA/S-95 suffered from extensive termination.

Table 5 Characteristics of poly(glycidyl methacrylate-ran-styrene) macroinitiators and styrene chain extensions at 110 °C in 50 wt% dioxane solution

Expt. ID	Macroinitiator			Chain extended product
	M n/kg mol^−1	M w/Mn	F GMA	M n/kg mol^−1	M w/Mn	F GMA
a The Mn and Mw/Mn compositions and FGMA are given for the fractionated product.
GMA/S-10-S	8.9	1.25	0.05	52.4	1.33	0.01
GMA/S-50-S	8.6	1.32	0.32	47.6	1.44	0.08
GMA/S-95-S	11.4	1.38	0.92	42.9^a	1.61	0.13

---

Fig. 6 Gel permeation chromatograms (GPC) for chain extensions from various poly(GMA-ran-S) macroinitiators with a second batch of styrene. All chain extensions were done in 50 wt% dioxane solutions at 110 °C. The chain extensions shown are: (a) GMA/S-10-S (dashed line is the GMA/S-10 macroinitiator, solid line is the block copolymer); (b) GMA/S-50-S (dashed line is the GMA/S-50 macroinitiator, solid line is the block copolymer); (c) GMA/S-95-S (dashed line is the GMA/S-95 macroinitiator, thin solid line is the block copolymer, thick solid line is the block copolymer after fractionation). The properties of the macroinitiator and the chain-extended species are summarized in Table 5.

It should be noted that for the macroinitiators used for the chain extension, the GMA content was not the only difference, as the macroinitiators were polymerized to different conversions. For instance, for the GMA/S-10 and GMA/S-50 macroinitiators, the conversions were similar (0.25 and 0.33, respectively) resulting in a smooth shift to the block copolymer. However, for the GMA/S-95-S chain extension experiment, the conversion was higher (0.59) and there is an increased likelihood of irreversible termination reactions, resulting in more inactive macroinitiator chains. This was easily seen in the GPC chromatogram of GMA/S-95-S as the peak of the chain extended copolymer was clearly overlapping with the peak due to the macroinitiator before chain extension (Fig. 6c). Fractionation of the crude block copolymer removed much of the dead macroinitiator and a product with a much lower polydispersity (M_n = 42.3 kg mol⁻¹, M_w/M_n = 1.61, F_GMA = 0.13) was recovered. Nevertheless, the chain extension experiments were promising, as block copolymers could be accessible from a wide range of macroinitiator compositions.

Conclusions

In this study, GMA and styrene were copolymerized in 1,4-dioxane solution at 90 °C with NHS-BlocBuilder initiator. Apparent rate constants 〈k_p〉[P˙] increased substantially for higher GMA feed compositions (f_GMA,0 > 80 mol%). Chain growth (M_nversus conversion) during the copolymerizations matched the theoretical prediction relatively well at low conversion <40% but started deviating at higher conversions. M_w/M_n steadily decreased during the copolymerization and levelled to about 1.3–1.4 in almost all cases. The macroinitiators with relatively low GMA compositions were substantially “living” as nearly monomodal shifts of M_n were observed after chain extension with a fresh batch of styrene monomer. The only exception was for the experiment that had high GMA content and macroinitiator conversion (experiment GMA/S-95-S), where a substantial fraction of dead macroinitiator chains was present.

NHS-BlocBuilder was successful in controlling the GMA/S copolymerizations and unlike the copolymerizations done with neat BlocBuilder, side-reactions were sufficiently suppressed to prevent branching/cross-linking that lead to insoluble product. Furthermore, the GMA/S copolymerizations were done without the additional SG1 free nitroxide that is normally required for methacrylate polymerizations with BlocBuilder. Sufficient free nitroxide was released by the NHS-BlocBuilder to sufficiently control the polymerizations up to very high GMA feed loadings. This greatly simplifies the formulation and also allows further post-polymerization transformations of the NHS end group. This marks the first time that GMA was successfully incorporated into resins at high GMA loadings with NMP over a wide composition range.

Acknowledgements

This work was supported by TEAMEC Energy Research. The authors thank Scott Schmidt and Noah Macy of Arkema, Inc. for their aid in obtaining the BlocBuilder alkoxyamine initiator used for this work. We thank Prof. Bernadette Charleux (CPE Lyon) for fruitful discussions about GMA polymerizations with BlocBuilder.

References

1. H. Fischer, Macromolecules, 1997, 30, 5666–5672.
2. H. Fischer, J. Polym. Sci., Part A: Polym. Chem., 1999, 37, 1885–1901.
3. H. Fischer, Chem. Rev., 2001, 101, 3581–3610.
4. K. Matyjaszewski and J. Xia, Chem. Rev., 2001, 101, 2921–2990.
5. M. Kamigaito, T. Ando and M. Sawamoto, Chem. Rev., 2001, 101, 3689–3746.
6. J.-S. Wang and K. Matyjaszewski, J. Am. Chem. Soc., 1995, 117, 5614–5615.
7. C. J. Hawker, G. Barclay, A. Orellana, J. Dao and W. Devonport, Macromolecules, 1996, 29, 5245–5254.
8. C. J. Hawker, A. W. Bosman and E. Harth, Chem. Rev., 2001, 101, 3661–3688.
9. G. Moad, J. Chiefari, Y. K. Chong, J. Krstina, R. T. A. Mayadunne, A. Postma, E. Rizzardo and S. H. Thang, Polym. Int., 2000, 49, 993–1001.
10. G. Moad, R. T. A. Mayadunne, E. Rizzardo, M. Skidmore and S. H. Thang, Macromol. Symp., 2003, 192, 1–12.
11. A. Oyane, T. Ishizone, M. Uchida, K. Furukawa, T. Ushida and H. Yokoyama, Adv. Mater., 2005, 17, 2329–2332.
12. D. C. Popescu, R. Lems, N. A. A. Rossi, C. T. Yeoh, J. Loos, S. J. Holder, C. V. C. Bouten and N. A. J. M. Sommerdijk, Adv. Mater., 2005, 17, 2324–2329.
13. S. P. Cullen, S. Ha, M. G. Lagally and P. Gopalan, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 5826–5838.
14. A. S. Zalusky, R. Olayo-Valles, J. H. Wolf and M. A. Hillmyer, J. Am. Chem. Soc., 2002, 124, 12761–12773.
15. F. Guo, K. Jankova, L. Schulte, M. Vigild and S. Ndoni, Macromolecules, 2008, 41, 1486–1493.
16. S. Yang, I. Ryu, H. Kim, J. Kim, S. Jang and T. P. Russell, Adv. Mater., 2006, 18, 709–712.
17. J. Rzayev and M. A. Hillmyer, Macromolecules, 2005, 38, 3–5.
18. S. Park, J. Wang, B. Kim, J. Xu and T. P. Russell, ACS Nano, 2008, 2, 766–772.
19. E. A. Jackson and M. A. Hillmyer, ACS Nano, 2010, 4, 3548–3553.
20. L. Chen and M. A. Hillmyer, Macromolecules, 2009, 42, 4237–4243.
21. S. A. F. Bon, M. Bosveld, B. Klumperman and A. L. German, Macromolecules, 1997, 30, 324–326.
22. D. Benoit, S. Grimaldi, S. Robin, J.-P. Finet, P. Tordo and Y. Gnanou, J. Am. Chem. Soc., 2000, 122, 5929–5939.
23. S. Grimaldi, J.-P. Finet, F. O. Le Moigne, A. Zeghdaoui, P. Tordo, D. Benoit, M. Fontanille and Y. Gnanou, Macromolecules, 2000, 33, 1141–1147.
24. D. Benoit, V. Chaplinski, R. Braslau and C. J. Hawker, J. Am. Chem. Soc., 1999, 121, 3904–3920.
25. B. Grassl, G. Clisson, A. Khoukh and L. Billon, Eur. Polym. J., 2008, 44, 50–58.
26. H. Götz, E. Harth, S. M. Schiller, C. W. Frank, W. Knoll and C. J. Hawker, J. Polym. Sci., Part A: Polym. Chem., 2002, 40, 3379–3391.
27. B. Charleux, J. Nicolas and O. Guerret, Macromolecules, 2005, 38, 5485–5492.
28. G. S. Ananchenko, M. Souaille, H. Fischer, C. Le Mercier and P. Tordo, J. Polym. Sci., Part A: Polym. Chem., 2002, 40, 3264–3283.
29. C. Dire, J. Belleney, J. Nicolas, D. Bertin, S. Magnet and B. Charleux, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 6333–6345.
30. C. Burguière, M.-A. Dourges, B. Charleux and J.-P. Vairon, Macromolecules, 1999, 32, 3883–3890.
31. R. McHale, F. Aldabbagh and P. B. Zetterlund, J. Polym. Sci., Part A: Polym. Chem., 2007, 45, 2194–2203.
32. Y. Guillaneuf, D. Gigmes, S. R. A. Marque, P. Tordo and D. Bertin, Macromol. Chem. Phys., 2006, 207, 1278–1288.
33. A. C. Greene and R. B. Grubbs, Macromolecules, 2010, 43, 10320–10325.
34. Y. Guillaneuf, D. Gigmes, S. R. A. Marque, P. Astolfi, L. Greci, P. Tordo and D. Bertin, Macromolecules, 2007, 40, 3108–3114.
35. C. Detrembleur, A. Mouithys-Mickalad, P. Teyssié and R. Jérôme, e-Polym., 2002, 004.
36. C. Detrembleur, P. Teyssié and R. Jérôme, Macromolecules, 2002, 35, 1611–1621.
37. V. Sciannamea, R. Jérôme and C. Detrembleur, Chem. Rev., 2008, 108, 1104–1126.
38. C. Detrembleur, V. Sciannamea, C. Koulic, M. Claes, M. Hoebeke and R. Jérôme, Macromolecules, 2002, 35, 7214–7223.
39. J. Vinas, N. Chagneux, D. Gigmes, T. Trimaille, A. Favier and D. Bertin, Polymer, 2008, 49, 3639–3647.
40. J. Nicolas, G. Mantovani and D. M. Haddleton, Macromol. Rapid Commun., 2007, 28, 1083–1111.
41. M. Schuster, C. Turecek, G. Weigel, R. Saf, J. Stampfl, F. Varga and R. Liska, J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 7078–7089.
42. Z. Chen, H. Bao and J. Liu, J. Polym. Sci., Part A: Polym. Chem., 2001, 39, 3726–3732.
43. P. F. Cañamero, J. L. de la Fuente, E. L. Madruga and M. Fernández-García, Macromol. Chem. Phys., 2004, 205, 2221–2228.
44. G. G. Godwin, C. S. Jone Selvamalar, A. Penlidis and S. Nanjundan, React. Funct. Polym., 2004, 59, 197–209.
45. X. Zhang and H. Tanaka, J. Appl. Polym. Sci., 2001, 80, 334–339.
46. A. Brar and A. Goyal, Eur. Polym. J., 2008, 44, 4089–4098.
47. H. Yin, H. Zheng, L. Lu, P. Liu and Y. Cai, J. Polym. Sci., Part A: Polym. Chem., 2007, 45, 5091–5102.
48. R. B. Grubbs, J. M. Dean, M. E. Broz and F. S. Bates, Macromolecules, 2000, 33, 9522–9534.
49. B. Lessard and M. Maric, J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 2574–2588.
50. L. A. O'Neill and C. P. Cole, J. Appl. Chem., 1956, 6, 356–364.
51. P. Guegan, C. W. Macosko, T. Ishizone, A. Hirao and S. Nakahama, Macromolecules, 1994, 27, 4993–4997.
52. M. E. Stewart, S. E. George, R. L. Miller and D. R. Paul, Polym. Eng. Sci., 1993, 33, 675–685.
53. A. Moayeri, M. Eng. thesis, McGill University, 2011.
54. W. Wang and R. A. Hutchinson, Macromolecules, 2008, 41, 9011–9018.
55. K. Liang and R. A. Hutchinson, Macromolecules, 2010, 43, 6311–6320.
56. S. Beuermann and M. Buback, Prog. Polym. Sci., 2002, 27, 191–254.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c1py00190f

---