Synthesis of well-defined primary amine-based homopolymers and block copolymers and their Michael addition reactions with acrylates and acrylamides

Abstract

A series of well-defined primary amine-based AB diblock copolymers were synthesised via atom transfer radical polymerisation (ATRP) using 2-aminoethyl methacrylate hydrochloride (AMA), with the other block comprising the following comonomers: 2-(diisopropylamino)ethyl methacrylate (DPA), 2-hydroxypropyl methacrylate (HPMA) and 2-(methacryloyloxy)ethyl phosphorylcholine (MPC). These copolymers were prepared with reasonably narrow polydispersities (M_w/M_n ≈ 1.1–1.4) in either 80 : 20 or 95 : 5 2-propanol–water mixtures at 50 °C using a 2-(N-morpholino)ethyl isobutyryl bromide initiator. The chain extension efficiency of PAMA was also investigated using a further charge of AMA. Unfortunately, such ‘self-blocking’ was problematic due to both catalyst deactivation and termination of the living chain ends. Nevertheless, low polydispersity all-methacrylic diblock copolymers were obtained in high yields via sequential monomer addition, provided that AMA was used as the second monomer in such syntheses. PAMA₄₉ homopolymer and selected PAMA-based copolymers were reacted with various acrylates and acrylamides in aqueous solution at pH 9, with mean degrees of functionalisation being determined by ¹H NMR spectroscopy. This facile Michael addition chemistry provides access to a library of novel functional water-soluble homopolymers and diblock copolymers.

---

Introduction

Over the last decade or so, the emergence of living radical polymerisation techniques such as atom transfer radical polymerisation (ATRP)^1–4 and reversible addition–fragmentation transfer (RAFT)^5–9 polymerisation has resulted in a paradigm shift in the synthesis of functional polymers of controlled architectures. Thousands of papers have described facile syntheses of relatively well-defined vinyl polymers, including many examples based on water-soluble/hydrophilic monomers.^10–16 Nevertheless, there are still relatively few studies that have exploited ATRP or RAFT for the synthesis of controlled-structure primary amine-based vinyl polymers.^17–25

2-Aminoethyl methacrylate hydrochloride (AMA) is a commercially available primary amine-based methacrylic monomer. However, it is both relatively expensive and only available as a technical grade of ∼90% purity. Fortunately, AMA is readily prepared by reacting methacryloyl chloride with molten 2-aminoethanol hydrochloride at 95 °C.²⁶ It has been previously used as a comonomer for the preparation of cationic latexes^27,28 and has also been oligomerised using catalytic chain transfer polymerisation to afford polydisperse AMA-based macromonomers.²⁹ Narain et al. reported that AMA can be reacted with lactobionolactone in order to prepare a new cyclic sugar methacrylate.^21,26

Recently we demonstrated the successful polymerisation of AMA directly in its hydrochloride form via both RAFT and ATRP.²⁰ Linear PAMA homopolymers and AB diblock copolymers [where A = AMA and B = 2-(diisopropylamino)ethyl methacrylate [DPA]] were prepared with narrow polydispersities (M_w/M_n ≈ 1.2–1.3) at 70 °C using cumyl dithiobenzoate (CDB) as a RAFT chain transfer agent. In addition, ATRP was used to prepare a range of PAMA homopolymers in alcohol–water mixtures and also several examples of well-defined PAMA-based diblock copolymers using poly(ethylene oxide) macro-initiators. In its neutral form, AMA monomer is rapidly converted into 2-hydroxyethyl methacrylamide. However, this internal rearrangement can be completely suppressed by ensuring that the AMA monomer is maintained in its hydrochloride salt form. PAMA is much more stable than AMA monomer, but slow degradation occurs via at least two pathways in alkaline media.¹⁸

Michael addition chemistry benefits from mild reaction conditions, no by-products, high functional group tolerance, high conversions and favourable reaction rates.³⁰ Typical ‘Michael donors’ include amines, enolates, thiols and phosphines. These can react with a wide range of ‘Michael acceptors’, including acrylates, acrylamides, acrylonitriles, maleimides, methacrylates, cyanoacrylates and vinyl sulfones.³⁰ Michael addition has been exploited in step growth polymerisation to prepare biocompatible poly(amido amines),^31–33 poly(amino esters)³⁴ and poly(aspartamides).³⁵ There are also numerous literature examples of the modification of polymers using Michael addition. For example, it has also been used to conjugate synthetic polymers to proteins by reacting vinyl sulfone-terminated poly(ethylene oxide) chains with the primary amine groups on either lysine or cysteine groups to produce useful ‘pegylated’ proteins for therapeutic applications.³⁶ Michael addition has also been used to modify surfaces,³⁷ produce cross-linked networks,^38,39 gels⁴⁰ and prepare other types of bioconjugates^41–43 and, more recently, model branched copolymers.²²

Herein we report the synthesis of well-defined PAMA-based diblock copolymers and the attempted chain extension of PAMA homopolymers via sequential monomer addition. Moreover, post-polymerisation functionalisation of these homopolymers and PAMA-based diblock copolymers via Michael addition has been conducted in aqueous solution using various acrylates and acrylamides to yield a library of novel copolymers with identical molecular weight distributions (see Scheme 1).

Scheme 1 Schematic representation of the synthesis of all-methacrylic PAMA-based diblock copolymers by using either 2-(diisopropylamino)ethyl methacrylate (DPA), 2-(methacryloyloxy)ethyl phosphorylcholine (MPC) or 2-hydroxypropyl methacrylate (HPMA) in the first-stage polymerisation.

Experimental

Materials

2-Hydroxypropyl methacrylate (HPMA, >97%) was donated by Cognis Performance Chemicals Ltd (Hythe, UK). 2-(Methacryloyloxy)ethyl phosphorylcholine (MPC, 99.9%) was donated by Biocompatibles, UK. These two monomers were used without further purification. 2-(Diisopropylamino)ethyl methacrylate (DPA) was purchased from Scientific Polymer Products (USA); inhibitor was removed by passing through a basic alumina column prior to use. Methacryloyl chloride (>97%), ethanolamine hydrochloride (>98%), hydroquinone (99%), Cu(I)Cl (99.995%), 2,2′-bipyridine (bpy, 99%), N,N,N′,N″,N″-pentamethyldiethylenetriamine (99%), 4-(2-hydroxyethyl)morpholine, poly(ethylene glycol) methyl ether acrylate (PEGA, 99%), 2-(dimethylamino)ethyl acrylate (DMAA, 98%), N,N-dimethylacrylamide (DMAc, 99%), sodium acrylamido-2-methyl-1-propanesulfonate (AMPS, 50 w/w% in water), (3-acrylamidopropyl)trimethylammonium chloride (ATAC, 75 w/w% in water), 4-acryloylmorpholine (AM, 97%), sodium acrylate (NaAA, 99%), acrylamidoglycolic acid monohydrate (AGA, 96%), sodium vinyl sulfonate (SVS, 98%), methyl acrylate (MA, 99%), D₂O, NaOD and DCl were all purchased from Aldrich and were used as received. Methanol and 2-propanol (IPA) were purchased from Fisher and were used as received. De-ionised water was used in all experiments involving aqueous or mixed aqueous solvents. NMR solvents (CD₃OD and CDCl₃) and regenerated cellulose dialysis membrane (Spectra/Por 6, molecular weight cut-off 1000) were all purchased from Aldrich and were used as received.

AMA monomer synthesis

AMA monomer was synthesised using a previously reported protocol.²⁶ 2-Aminoethanol hydrochloride (65 g, 0.68 mol), methacryloyl chloride (100 mL, 0.96 mol), and hydroquinone (0.50 g) were added to a three-necked round-bottomed flask fitted with a condenser. The molten 2-aminoethanol hydrochloride salt was esterified in the melt at 90–95 °C under a nitrogen atmosphere for 1 h and the reaction was maintained at 70–75 °C for a further 2 h. The HCl gas evolved during this reaction was neutralised using an aqueous NaOH solution connected to the reaction vessel. The crude product was cooled to 40 °C and diluted with THF (50 mL) to form a slurry. This slurry was then precipitated into excess n-pentane (500 mL). The resulting creamy white precipitate was isolated by filtration, washed thoroughly with n-pentane (500 mL), and dried under vacuum. It was then recrystallised using a 7 : 3 ethyl acetate–IPA mixture at −25 °C. After recrystallisation, the crude product was washed on a frit with a minimum amount of THF. The final product was dried under vacuum to produce an overall yield of 66% (>99% purity as judged by ¹H NMR). The 2-(4-morpholino)ethyl 2-bromoisobutyrate (ME–Br) ATRP initiator, PEO-based macro-initiator (PEO_n–Br where n = 45) and a PPO-based macro-initiator (PPO₄₃–Br) were synthesised according to the literature.^12,44,45

Homopolymerisation of AMA

The homopolymerisation of AMA was performed using ME–Br initiator in 80 : 20 IPA–H₂O, as described previously.²⁰ In a typical ATRP synthesis, AMA (1.47 g, 9.0 mmol), ME–Br (50 mg, 0.18 mmol, target DP_n = 50) and bpy (56 mg, 0.36 mmol) were added to a round-bottomed flask. After sealing the flask with a rubber septum, three vacuum/nitrogen cycles were performed to remove oxygen. Degassed 80 : 20 IPA–H₂O (3.1 mL, such that [AMA]_o = 33 w/v%) was introduced into the flask and this stirred solution was immersed in an oil bath at 50 °C. On addition of Cu(I)Cl catalyst (18 mg, 0.18 mmol) the reaction solution turned dark brown, indicating the onset of polymerisation. After 6 h, the polymerising solution was diluted with water and dialysed against water (Spectrapor 6, molecular weight cut-off 1000 Da) for three days to remove the spent blue Cu(II) catalyst. During dialysis, the blue/green polymer solution gradually became colourless. The aqueous solution was then freeze-dried from water overnight to give a white polymer. The attempted ‘self-blocking’ chain extension of PAMA₅₀ involved essentially the same protocol: a second charge of AMA monomer was added at high conversion (as judged by ¹H NMR). In the case of the attempted diblock copolymer chain extension experiments, either MPC, HPMA or DPA was added in situ to PAMA homopolymer (see Table 1 for further synthesis conditions).

Table 1 Summary of conversion, molecular weight and polydispersity data for the attempted chain extension of PAMA and its diblock copolymerisation from PHPMA, PMPC or PDPA homopolymers via a one-pot ATRP protocol using (unless otherwise stated) ME–Br initiator at 50 °C, a fixed monomer concentration of 33 w/v% and [ME–Br] : [CuCl] : [bpy] relative molar ratios of 1 : 1 : 2

Target copolymer composition	1^st Block					2^nd Block
	Solvent	Time/h	Conversion^a (%)	M n	M w/Mn^b	Solvent	Time/h	Conversion^a (%)	M n	M w/Mn^b
a Conversions were determined using ^1H NMR spectroscopy by comparing the area of the vinyl proton resonances to the total methyl signal. b As determined by aqueous GPC at pH 3 using a series of poly(2-vinylpyridine) calibration standards. c Determined by DMF GPC after modification of AMA units with excess methyl acrylate in the presence of NEt3. d 50 w/v% monomer concentration.
PAMA30–PAMA30	80 : 20 IPA–H2O	3	96	7800	1.11	80 : 20 IPA–H2O	24	36	8200	1.14
PAMA30–PHPMA30	80 : 20 IPA–H2O	3	99	—	—	80 : 20 IPA–H2O	24	69	8600^c	1.40^c
PAMA30–PDPA30	80 : 20 IPA–H2O	2.5	98	6700	1.19	80 : 20 IPA–H2O	15.5	52	Bimodal
PHPMA30–PAMA30	80 : 20 IPA–H2O	5.5	98	—	—	80 : 20 IPA–H2O	24	95	16 200^c	1.40^c
PMPC30–PAMA30	Methanol, 20 °C^d	1.3	100	8700	1.12	80 : 20 IPA–H2O	24	99	11 200	1.11
PDPA30–PAMA30	IPA^d	5	99	9900	1.29	80 : 20 IPA–H2O	24	93	11 900	1.28
PDPA50–PAMA30	IPA^d	7.5	98	12 100	1.21	80 : 20 IPA–H2O	15.5	96	13 200	1.28

---

AMA polymerisation using macro-initiators

Homopolymerisation of AMA was also performed in turn using two ATRP macro-initiators, PEO₄₃–Br and PPO₄₅–Br in either 95 : 5 IPA–H₂O or 80 : 20 IPA–H₂O mixtures. The protocol for each AMA polymerisation was essentially the same, apart from the choice of initiator and solvent. In a typical ATRP synthesis, PEO₄₅–Br (1.00 g, 0.47 mmol) and bpy (147 mg, 0.94 mmol) were added to a flask under a nitrogen atmosphere. Three vacuum/nitrogen cycles were performed to remove oxygen before a deoxygenated solution of AMA (2.31 g, 14.0 mmol) in 95 : 5 IPA–H₂O (6.6 mL, 33%) was added. On addition of Cu(I)Cl catalyst (46 mg, 0.47 mmol) the reaction solution turned dark brown, indicating the onset of polymerisation. After 6 h stirring at 50 °C, the polymerising solution was diluted with water and dialysed against water (molecular weight cut-off 1000 Da) for three days to remove the Cu(II) catalyst. During dialysis, the blue/green copolymer solution gradually became colourless. The aqueous copolymer solution was then freeze-dried from water.

Chain extension and triblock copolymer syntheses were attempted using essentially the same protocol as that used for the homopolymerisation of AMA using macro-initiators. Thus, a second charge of AMA was added at high conversion (as judged by ¹H NMR) for self-blocking experiments (e.g. targeting PEO–PAMA–PAMA) and a new charge of DPA was employed for the PEO–PAMA–PDPA triblock copolymer syntheses. A typical attempted synthesis of PEO₄₅–PAMA₃₀–PDPA₅₀ was as follows. In a 25 mL flask, PEO₄₅–Br (0.50 g, 0.23 mmol), bpy (74 mg, 0.46 mmol) and AMA (1.16 g, 7.0 mmol) were added and sealed with a rubber septum. Three vacuum/nitrogen cycles were performed to remove the oxygen before degassed IPA (3.3 mL, 33%) followed by CuCl (23 mg, 0.23 mmol) were introduced into the flask. After 2.5 h, the conversion was greater than 95% and a deoxygenated solution of DPA (2.49 g, 11.6 mmol) in 80 : 20 IPA–H₂O (2.5 mL, 50 w/v%) was added under a positive nitrogen pressure. Purification was carried out using the same protocol as that for the PEO–PAMA diblock copolymers.

Synthesis of all-methacrylic PAMA-based diblock copolymers

A typical synthesis of PDPA₃₀–PAMA₃₀ was as follows. In a 25 mL flask, ME–Br (50 mg, 0.18 mmol), bpy (56 mg, 0.36 mmol) and DPA (1.143 g, 5.4 mmol) were added and sealed with a rubber septum. Three vacuum/nitrogen cycles were performed to remove oxygen before degassed IPA (1.2 mL, 50 w/v%) followed by CuCl (18 mg, 0.18 mmol) were introduced into the flask. After 5 h the conversion was greater than 95% and a deoxygenated solution of AMA (0.89 g, 5.4 mmol), in 80 : 20 IPA–H₂O (1.8 mL, 33 w/v %), was added under a positive nitrogen pressure. After 24 h stirring at 50 °C, the polymerising solution was diluted with water and dialysed against water (molecular weight cut-off 1000 Da) for three days to remove the Cu(II) catalyst. During dialysis, the blue/green copolymer solution gradually became colourless. The aqueous copolymer solution was then freeze-dried from water. AMA was also added as the second block to MPC and HPMA homopolymerisations.

Michael addition to PAMA homopolymer

In a typical Michael addition, PAMA₄₉ (0.04 g, 0.24 mmol of AMA units) was dissolved in 0.80 g of D₂O to produce a 5.0 w/w% solution. To this solution, PEGA was added (0.12 g, 0.26 mmol, 1.1 molar equivalents based on AMA). The pH of this PAMA₄₉ solution was adjusted to pH 9 using NaOD. The extent of PAMA₄₉ modification was monitored by ¹H NMR for 6 h at 50 °C. Reactions were also carried out on a larger scale in H₂O. Typically, PAMA₄₉ (0.50 g, 3.0 mmol of AMA units) was dissolved in 10.0 g H₂O to produce a 5.0 w/w% solution. To this solution, DMAA was added (4.62 g, 3.30 molar equivalents based on AMA). The pH of this PAMA₄₉ solution was adjusted to pH 9 using NaOH. After stirring for 24 h at 50 °C, solutions were dialysed against water to remove any excess unreacted acrylate, followed by freeze-drying. The overall extent of the modification was calculated using ¹H NMR from a purified sample.

Characterisation techniques

¹H NMR spectroscopy. All ¹H NMR spectra were recorded in either D₂O with NaOD/HCl, CDCl₃, CD₃OD or d₅-pyridine using a 250 or 400 MHz Bruker AV250/AV400 spectrometer.

Aqueous GPC. The molecular weights and polydispersities of all the PAMA homopolymers, the PDPA–PAMA, PMPC–PAMA diblock copolymers and PEO–PAMA–PAMA triblock copolymers, were determined by aqueous GPC at 35 °C using a PL Aquagel-OH 40 and a Aquagel-OH 30 column connected in series to a Polymer Labs ERC-7517A refractive index detector. The eluent was a pH 3.3 buffer solution comprising 0.30 M NaH₂PO₄ and 1.0 M acetic acid at a flow rate of 1.0 mL min⁻¹. Eight near-monodisperse poly(2-vinyl pyridine) standards (M_p = 1480–117 000 g mol⁻¹) were used for calibration. Data were analysed using PL Cirrus GPC software (version 2.0) supplied by Polymer Laboratories.

DMF GPC. The molecular weights and polydispersities of the PPO₄₃–PAMA₆₀ and PHPMA₃₀–PAMA₃₀ diblock copolymers were determined by DMF GPC at 70 °C. The GPC set-up comprised three Polymer Laboratories PL gel 10 µm Mixed-B columns in series with a Viscotek TriSEC Model 302 refractive detector. The flow rate was 1.0 mL min⁻¹ and the mobile phase contained 10 mmol LiBr. Ten near-monodisperse PMMA homopolymer standards (M_p = 2000–300 000 g mol⁻¹) were used for calibration. The data were analysed using Viscotek TriSEC 3.0 software. Prior to analysis, the AMA groups on these copolymers were derivatised with excess methyl acrylate in DMF containing triethylamine at 20 °C for 24 h.

Results and discussion

We have previously reported the homopolymerisation of AMA in its HCl salt form via ATRP.²⁰ Various reaction conditions were investigated, including polymerisation temperature and solvent composition. Empirically, it was found that very high monomer conversions and relatively narrow molecular weight distributions (M_w/M_n < 1.25) were obtained within 3 h at 50 °C using an 80 : 20 IPA–water mixture and an [AMA]_o of 33 w/v%. These optimised conditions were also employed in the present work for the first-stage polymerisation of AMA, with very similar results being obtained (see entries 1–3 in Table 1). However, one aspect not investigated in our earlier study was the living character of the PAMA chains at the end of such polymerisations, i.e. to what extent can further chain growth be achieved by the addition of a second monomer charge? This important question is addressed in the present work. Chain extension experiments were conducted to assess the living character of PAMA homopolymers synthesised via ATRP. For example, a second AMA monomer charge was added to a PAMA₃₀ synthesis at 96% AMA conversion (see entry 1 in Table 1). However, only 36% monomer conversion was achieved for the second-stage polymerisation after 24 h. As a result, only a relatively small increase in molecular weight was observed by GPC. Two further chain extension experiments were also performed using HPMA and DPA monomer, respectively. Somewhat higher conversions were achieved (52% for DPA and 69% for HPMA) (see entries 2 and 3 in Table 1). However, the GPC trace for HPMA chain extension was broad (M_w/M_n = 1.40) and for the DPA chain extension the GPC trace was both broad and bimodal.

As previously reported, a series of PEO–PAMA diblock copolymers may be readily prepared by ATRP.²⁰ PEO₄₅–PAMA₃₀ proved to be fully soluble in the 80 : 20 IPA–H₂O reaction solution, but adding a second charge of AMA monomer (target DP = 30) to this reaction once the first-stage polymerisation had reached high conversion (96%) only led to 63% conversion for the second-stage polymerisation. This incomplete conversion was accompanied by only a relatively modest increase in molecular weight (M_n) from 6800 to 9300 as judged by aqueous GPC.

There are two possible explanations for the incomplete chain extension of PAMA in these ‘self-blocking’ experiments: (1) chain-end termination has occurred, resulting in removal of the terminal bromine and hence preventing further propagation; (2) the catalyst has become deactivated, possibly by interaction with the primary amine groups on PAMA or by reacting with the hydrochloric acid associated with each AMA residue. To investigate the second possibility further, the chain extension of PAMA was repeated but this time, on addition of the second monomer charge (this time at 98% PAMA conversion), further catalyst and ligand were added (doubling the overall catalyst content) in order to compensate for any catalyst deactivation during the first-stage polymerisation. In this case, the second PAMA block reached high conversion (∼100%) and the M_n increased from 4100 to 7900 indicating successful chain extension, see Fig. 1. This significant improvement in conversion seems to support the hypothesis of catalyst deactivation. However, there was also an increase in M_w/M_n from 1.12 to 1.40. This broadening of the molecular weight distribution was accompanied with the appearance of a low molecular weight shoulder in the GPC trace of the chain-extended homopolymer. This suggests that some premature chain-end termination also occurs in these ‘self-blocking’ experiments.

Fig. 1 Aqueous GPC traces obtained for the chain extension of PAMA: PAMA₃₀–PAMA₃₀ diblock copolymer (dashed line) and PAMA₃₀ (black line) prepared by ATRP.

Previously, we reported a macro-initiator approach to prepare PAMA-based diblock copolymers via ATRP²⁰ using two PEO-based macro-initiators with mean degrees of polymerisation (DP) of 45 and 113, respectively. Homopolymerisation of AMA (DP = 30) with either PEO₄₅–Br or PEO₁₁₃–Br at 50 °C in 95 : 5 IPA–H₂O proceeded to high conversion within 6 h, whereas a target DP of 60 required 15 h for complete conversion in an 80 : 20 IPA–H₂O mixture; these purified diblock copolymers had polydispersities of 1.26 or lower. A PPO₄₃–PAMA₆₀ diblock copolymer was successfully prepared using a 80 : 20 IPA–H₂O solvent mixture by ATRP. This solvent composition also gave reasonably good control (M_w/M_n = 1.30, M_n = 23 900) and high conversions (99%, within 48 h); it appears to be close to optimal for such syntheses.

Unfortunately, the attempted chain extension of PEO₄₅–PAMA₃₀ with a further AMA charge was also unsuccessful. A high conversion (>98%) was obtained for the second block and there was a corresponding increase in M_n from 3300 to 5900, see Fig. 2. Unfortunately chain-end termination was again observed, as judged by the appearance of a low molecular tail in the GPC trace for PEO₄₅–PAMA₃₀–PAMA₃₀ and the concomitant increase in M_w/M_n from 1.13 to 1.49. Thus it seems that PAMA chains prepared by ATRP under the stated conditions do not have sufficient living character to allow chain extension using other methacrylic monomers.

Fig. 2 Aqueous GPC traces obtained for the chain extension of PAMA: PEO₄₅–PAMA₃₀–PAMA₃₀ triblock copolymer (dashed line) and PEO₄₅–PAMA₃₀ diblock copolymer (black line) prepared by ATRP.

In contrast, reasonably well-defined all-methacrylic diblock copolymers can be synthesised in good yield via sequential monomer addition provided that AMA is added as the second monomer (see entries 4–6 in Table 1).⁴⁶

For example, the one-pot ATRP synthesis of PHPMA₃₀–PAMA₃₀ was conducted in 80 : 20 IPA–H₂O for both blocks. High conversions were achieved (>95%) for the first- and second-stage polymerisation but the diblock copolymer polydispersity by DMF GPC (after modification of the AMA residues with methyl acrylate; see later) was a little higher than expected (M_w/M_n = 1.40). However, for the analogous syntheses of the PMPC–PAMA and PDPA–PAMA diblocks, it proved necessary to polymerise the MPC and DPA monomers in methanol and IPA, respectively. Using 80 : 20 IPA–H₂O for the first block in the case of MPC led to solubility problems and hence some loss of control over the polymerisation. This is attributed to the well-known co-nonsolvency exhibited by PMPC in certain alcohol–water mixtures.^47–49

Once these first PMPC or PDPA blocks had reached high conversion (>98%), AMA monomer was added as a 33% solution in 80 : 20 IPA–H₂O. Final polydispersities were relatively narrow (M_w/M_n ≈ 1.11–1.28) and in all cases high conversions for the second block were achieved (>93%). An increase in molecular weight for the PAMA chain extension to give PDPA₃₀–PAMA₃₀ was confirmed by GPC, see Fig. 3.

Fig. 3 Aqueous GPC traces obtained for a PDPA₃₀–PAMA₃₀ diblock copolymer (dashed line; see entry 6 in Table 1) and the corresponding PDPA₃₀ homopolymer precursor (solid line) prepared by ATRP using sequential monomer addition (with the DPA monomer polymerised first).

Michael addition reactions

A series of Michael addition reactions were conducted using a near-monodisperse PAMA₄₉ homopolymer (M_n = 1.24; M_w/M_n = 9500). The pendent primary amine groups on this homopolymer were modified with either acrylates or acrylamides. Reactions were conducted in aqueous solution at pH 9. This particular pH was chosen for two reasons. Firstly, the pK_a of PAMA homopolymer is 7.6,²⁰ hence above this value the AMA groups will be mainly in their neutral amine form, which is a prerequisite for successful Michael addition. Secondly, PAMA homopolymer is stable for up to 48 h at pH 9 (5.0 w/w% solution at 50 °C), but degrades in more alkaline media.¹⁸

Michael addition between a primary amine and an acrylate is a well-known reaction in organic chemistry. Generally, polymer-analogous reactions tend to be slower than small molecule reactions and careful optimisation of the reaction conditions is usually required to ensure high conversions and to prevent side reactions.⁵⁰ The first two acrylates studied were poly(ethylene glycol) methyl ether acrylate (PEGA), a well-known hydrophilic monomer, and 2-(dimethylamino)ethyl acrylate (DMAA), a pH-responsive monomer. Their chemical structures are shown in Scheme 2a.

Scheme 2 Modification of PAMA polymers by Michael addition with (a) acrylates (e.g. poly(ethylene glycol) methyl ether acrylate (PEGA) or 2-(dimethylamino)ethyl acrylate (DMAA)) or (b) acrylamides (e.g.N,N-dimethylacrylamide (DMAc), sodium acrylamido-2-methyl-1-propanesulfonate (AMPS), 3-acrylamidopropyltrimethylammonium chloride (ATAC), 4-acryloylmorpholine (AM)).

Kinetic studies of the extent of acrylate addition were conducted with PEGA and DMAA. These two acrylates had very similar rates of reaction. In both cases essentially complete conversion was achieved within 5 h at 50 °C for 5.0 w/w% PAMA₄₉ using an acrylate–amine molar ratio of 1.1 : 1, see Fig. 4. These reactions were conveniently monitored by the disappearance of vinyl groups at 5.5–6.5 ppm by ¹H NMR in D₂O/NaOD.

Fig. 4 Conversion vs. time plots for the reaction of PAMA₄₉ with PEGA (triangles) and DMAA (squares) at relative molar ratios of [PEGA or DMAA] : [AMA] of 5 : 1 at 50 °C, pH 9 and 5.0 w/w% PAMA₄₉ in D₂O.

The same conditions mentioned above were used to synthesise PEGA-modified PAMA and DMAA-modified PAMA on a larger scale in H₂O (5.0 w/w%, 50 °C, 24 h, 1.1 : 1 acrylate–amine molar ratio). After the reaction, any unreacted acrylate was removed via dialysis against water followed by lyophilisation. Comparing the ¹H NMR spectra for PAMA and PEGA monomer with the PEGA-modified PAMA spectrum (Fig. 5) confirmed that Michael addition had occurred. Moreover, there was no sign of any vinyl signals (δ 5.5–7.0) from the monomer starting material after purification and new signals appeared (δ 2.6–3.4) from the two methylene protons, indicating formation of a β-aminoester. In addition, signals from the reacted PEGA were significantly broader compared to the monomer spectrum owing to their incorporation into polymer chains. The final apparent conversion of 103% for PEGA-modified PAMA₄₉ (see Table 2) was calculated by comparison of the four methylene protons labelled ‘c’ (δ 4.1–4.6) and ‘f’ (δ 2.5–2.9). This result may simply reflect the (in)accuracy of our NMR measurements, but it is also possible that a small fraction of PAMA units has reacted with two acrylates to give tertiary amine residues. For DMAA-modified PAMA, 100% conversion was confirmed by comparison of signals from overlapping peaks ‘c’ and ‘g’ (δ 3.6–3.0) with peaks ‘d,’ ‘e,’ ‘f’ and ‘h’ (δ 2.3–2.7).

Table 2 Summary of reaction conditions and conversions as judged by ¹H NMR for PAMA₄₉ reacted with either PEGA or DMAA. Conditions: acrylate–amine molar ratio = 1.1 : 1.0, 5.0 w/w% PAMA₄₉ solution at 50 °C for 24 h

Acrylate type	Total extent of reaction (%)	Degree of monosubstitution (%)	Degree of disubstitution (%)
PEGA	103	97	3
DMAA	100	100	0

---

Fig. 5 Assigned ¹H NMR spectra (D₂O) for (a) PAMA₄₉; (b) PEGA; (c) PEGA-modified PAMA₄₉; (d) DMAA; (e) DMAA-modified PAMA₄₉.

One of the main problems with the analysis of cationic polymers is their unsuitability for most GPC protocols. PAMA homopolymers can only be analysed by aqueous GPC using an acidic buffer (pH 2–3). In other solvents such as THF or DMF, PAMA is insoluble. This means that the analysis of PAMA-based block copolymers is problematic when the comonomer(s) are unsuitable for acidic aqueous GPC (e.g. HPMA). In principle, protection of the PAMA block with PEGA should aid the solubility of the block copolymer. Surprisingly, PEGA–PAMA₄₉ was only partially soluble in THF and DMF and was thus unsuitable for GPC analysis. In contrast, the PPO₄₃–PAMA₅₉ and PHPMA₂₉–PAMA₂₉ diblock copolymers were successfully derivatised directly in DMF by reaction with excess methyl acrylate. Triethylamine was added to the copolymer solution (10 g L⁻¹) until complete dissolution occurred. This solution was stirred at 20 °C for 24 h and filtered prior to DMF GPC analysis. Monomodal GPC curves were obtained with M_n values ranging from 5000 to 15 000 (vs. PMMA standards) and reasonably low polydispersities (M_w/M_n ∼ 1.3 to 1.4) in both cases. This indicates that derivatisation via Michael addition does not have any significant detrimental effect on the molecular weight distribution of the near-monodisperse precursor.

Four different acrylamides, DMAc, AMPS, ATAC and AM were selected for reaction with PAMA, see Scheme 2b. These particular monomers were selected for their known water solubility. First, the reaction between PAMA₄₉ and DMAc was attempted at an acrylamide–amine molar ratio of 5 : 1 (50 °C, pH 9, 5.0 w/w% PAMA₄₉). Even using this relatively large excess, it was found that the Michael addition of acrylamides with PAMA₄₉ was much slower than that observed for acrylates. This has been attributed to delocalisation of the amide nitrogen lone pair, which renders acrylamides less reactive.⁵¹ Nevertheless, the Michael addition of DMAc proceeded to 96% conversion after 48 h, see Fig. 6. Under the same conditions, the other three acrylamides (AMPS, ATAC and AM) proved to be somewhat faster, requiring less than 10 h for >95% conversion.

Fig. 6 Conversion vs. time plots for the reaction of PAMA₄₉ with DMAc, AMPS, ATAC and AM at a 5 : 1 acrylamide–amine molar ratio at 50 °C. Conditions: 5.0 w/w% PAMA₄₉ in D₂O/NaOD at pH 9.

Reactions were also conducted at acrylamide–amine molar ratios of 3 : 1 and 1 : 1 for the reaction of AMPS with PAMA₄₉ (50 °C, pH 9, 2.0 w/w%). These less forcing conditions resulted in slower reactions, as expected (see Fig. 7). The 1 : 1 reaction was significantly slower, requiring approximately one week to achieve ∼100% conversion, while the 3 : 1 and 5 : 1 reactions led to more than 95% conversion within 8.4 h and 4.8 h, respectively. Unsurprisingly, a higher PAMA₄₉ concentration (5.0 w/w%) reduced the reaction time for 100% amine conversion to 2.1 h (50 °C, pH 9, using a 5 : 1 AMPS–amine molar ratio), see Fig. 8. Moreover, ¹H NMR confirmed that this particular reaction continued to well beyond 100% conversion, producing essentially complete disubstitution of the AMA residues within 6 h, see Fig. 9.

Fig. 7 Conversion vs. time plots for the reaction of PAMA₄₉ with AMPS at acrylamide–amine molar ratios of 5 : 1 (diamonds), 3 : 1 (triangles) and 1 : 1 (squares) at 50 °C. Conditions: 2.0 w/w% PAMA₄₉ in D₂O/NaOD at pH 9 (inset shows the same data for the 5 : 1 and 3 : 1 molar ratios over shorter time scales).

Fig. 8 Conversion vs. time plots for the addition of AMPS to 2.0 w/w% and 5.0 w/w% PAMA₄₉ in D₂O/NaOD at pH 9 and 50 °C using an AMPS–amine relative molar ratio of 5 : 1.

Fig. 9 Conversion vs. time plot for the addition of AMPS to a 5.0 w/w% PAMA₄₉ solution in D₂O/NaOD at pH 9 and 50 °C using an AMPS–amine relative molar ratio of 5 : 1.

¹H NMR studies (conducted in D₂O) provided a convenient method for monitoring and optimising the modification of PAMA₄₉ with acrylamides. These conditions were successfully replicated in H₂O on a larger scale so as to synthesise DMAc-, AMPS-, ATAC- and AM-modified PAMA₄₉. Each acrylamide was added to a 5.0 w/w% solution of PAMA₄₉ (acrylamide–amine molar ratio = 5 : 1) at pH 9 and stirred at 50 °C for 48 h. After Michael addition, any unreacted acrylamide was removed by dialysis against water, followed by lyophilisation. All four reactions led to high conversions with varying degrees of disubstitution (125–196%), see Table 3 and Fig. 10.

Table 3 Summary of reaction conditions and conversions from ¹H NMR for PAMA₄₉ after reaction with DMAc, AMPS, ATAC or AM for 48 h at 50 °C and pH 9 and subsequent purification. Conditions: acrylate–amine molar ratio = 5 : 1, 5.0 w/w % PAMA₄₉ in D₂O

Acrylamide type	Total conversion (%)	Percentage of monosubstituted AMA repeat units (%)	Percentage of disubstituted AMA repeat units (%)
DMAc	190	10	90
AMPS	157	43	57
ATAC	125	75	25
AM	196	4	96

---

Fig. 10 Assigned ¹H NMR (D₂O) spectra for (a) PAMA₄₉; (b) DMAc; (c) DMAc-modified PAMA₄₉; (d) AMPS; (e) AMPS-modified PAMA₄₉; (f) ATAC; (g) ATAC-modified PAMA₄₉; (h) AM; (i) AM-modified PAMA₄₉ in D₂O.

It is worth emphasising that PAMA₄₉ is stable with respect to hydrolytic degradation under the chosen reaction conditions. Our earlier study¹⁸ showed that negligible chemical degradation of PAMA₄₉ (<1%) occurred after 48 h (50 °C, pH 9, 5.0 w/w% PAMA₄₉). Moreover, there is no evidence for any PAMA degradation in any of the final ¹H NMR spectra. This suggests that the reaction with acrylamides (or acrylates) is favoured over any degradation pathway and that the resulting derivatised polymer is much more stable than the PAMA precursor.

The reaction of PAMA₄₉ with acrylamidoglycolic acid (AGA) and sodium vinyl sulfonate (SVS) was also attempted. AGA appeared to react slowly; after five days (50 °C, pH 9, 5.0 w/w% and an acrylamide–amine ratio of 5 : 1) approximately 67% conversion was achieved, as judged by ¹H NMR. However, at the same time two new sharp peaks appeared in the spectrum and a control experiment conducted in the absence of any PAMA₄₉ confirmed that AGA degradation occurred on the same time scale, under the same conditions. It is quite likely that this amino acid-type monomer would actually react with PAMA₄₉ under different solvent/temperature conditions where monomer degradation either does not occur at all, or can be minimised. However, such optimisation is beyond the scope of the present study

SVS was selected in view of the facile Michael addition cross-linking chemistry of divinyl sulfone (DVS).⁵² However, this monofunctional analogue only reacted with PAMA₄₉ very slowly at a SVS–amine molar ratio of 5 : 1 at 50 °C. Thus, for a 5.0 w/w% solution of PAMA₄₉ the reaction reached only 47% conversion after 5 days. Moreover, leaving the reaction for two weeks only increased the conversion by a few percent. Because high conversions could not be achieved within a reasonable time scale, Michael addition derivatisation using this monomer was not pursued further.

AMPS was also utilised to modify selected PAMA-based diblock copolymers. Reaction with this reagent resulted in the synthesis of several novel anionic diblock copolymers with potentially interesting aqueous solution properties. The three diblock precursors utilised in these experiments were the pH-responsive PDPA₃₀–PAMA₂₈ and two thermo-responsive block copolymers, PHPMA₂₉–PAMA₂₉ and PPO₄₃–PAMA₅₉.

Michael addition was conducted using the same conditions as those employed for PAMA₄₉ homopolymer. An AMPS–amine molar ratio of 5 : 1 was added to a 5.0 w/w% copolymer solution and stirred for 48 h at 50 °C. Under these conditions, an AMPS conversion of 62% was calculated for PDPA₃₀–PAMA₂₈ using ¹H NMR by comparing the ‘c’ and ‘g’ signals at δ 4.0–4.5 (due to the PDPA and PAMA methylene protons) to the ‘p’ and ‘n’ signals at δ 2.5–3.0 (due to the methylene protons of the AMPS), see Table 4. The pH-induced micellisation behaviour of PDPA–PAMA has already been reported,²⁰ with well-defined PDPA-core micelles being formed above pH 6.3. We speculate that the micellar nature of the PDPA–PAMA diblock precursor at pH 9 most likely has an adverse effect on the rate of Michael addition of the AMPS with the AMA residues, since such core-shell nanostructures are likely to resist the build-up of anionic charge in the micelle coronas.

Table 4 Summary of reaction conditions and conversions for PDPA₃₀–PAMA₂₈, PHPMA₂₉–PAMA₂₉ and PPO₄₃–PAMA₅₉ after reaction with AMPS in water for 48 h at 50 °C. Conditions: 5 : 1 AMPS–amine molar ratio; 5.0 w/w% copolymer solution at pH 9

Diblock copolymer type	Time/h	Total conversion (%)	Percentage of monosubstituted AMA repeat units (%)	Percentage of disubstituted AMA repeat units (%)
PDPA30–PAMA28	48	62	62	0
PHPMA29–PAMA29	48	157	43	57
PPO43–PAMA59	48	93	93	0

---

In contrast, an AMPS conversion of 93% was calculated for PPO₄₃–PAMA₅₉ from the ¹H NMR spectrum of the derivatised copolymer after purification by dialysis. This is a relatively high conversion for monosubstituted amines and may well reflect the much less compact, hydrated nature of the PPO-core micelles compared to the PDPA–PAMA micelles, since this reduced structural order is less likely to resist the build-up of anionic charge density. It is well-known that PPO is a thermo-responsive polymer that displays LCST behaviour with an estimated cloud point of 15 °C.¹¹ Save et al. synthesised PPO₄₃–PGMA₇₀ (where PGMA = poly(glycerol monomethacrylate)) by ATRP and reported that this copolymer was molecularly dissolved at 5 °C while colloidal aggregates of 100–330 nm diameter were formed at 60 °C, as judged by DLS. In the present work, the PPO₄₃–PAMA₅₉ solution remained turbid throughout its reaction with the AMPS. It is likely that any reaction took place between the AMPS and the coronal PAMA₅₉ chains of PPO₄₃–core micelles, but further work is required to confirm this hypothesis.

The AMPS derivatisation of PHPMA₂₉–PAMA₂₉ was more encouraging, with a conversion of 157% being calculated from the ¹H NMR spectrum. This suggests a mixture of mono- and disubstituted amines in this particular case. The diblock precursor solution was also turbid at the Michael addition reaction temperature of 50 °C due to the thermo-responsive nature of the PHPMA block.⁵³ However, this appeared to have little effect on the rate of Michael addition; the overall conversion was comparable to that obtained for the PAMA₄₉ homopolymer under the same conditions.

Conclusions

Well-defined PAMA-based diblock copolymers were successfully prepared via ATRP in alcohol–water mixtures at 50 °C using sequential monomer addition. However, AMA monomer must be polymerised second in order to ensure good blocking efficiencies, high overall conversions and reasonable control (i.e. low polydispersities). Attempted chain extension of PAMA homopolymer was unsuccessful, apparently due to both catalyst deactivation and also premature chain termination. The former problem presumably involves irreversible oxidation of Cu(I) to Cu(II) and can be at least partially overcome by addition of fresh catalyst just prior to the second charge of AMA monomer. However, the problem of premature chain termination could not be alleviated and this results in significant contamination of the final chain-extended polymer with the homopolymer precursor.

PAMA homopolymer has been successfully reacted with a range of acrylates and acrylamides to obtain a library of new water-soluble polymers with the same relatively narrow molecular weight distribution. Acrylates proved to be more reactive than acrylamides, as expected. The former only require approximately stoichiometric molar ratios to form monosubstituted β-aminoesters. In contrast, the latter typically require excess reagents and more vigorous conditions, which often resulted in disubstituted primary amines.

Two reagents (sodium vinyl sulfonate and acrylamidoglycolic acid) proved to be much less successful, since they typically required extremely long reaction times and only gave incomplete conversions. Moreover, undesirable side reactions such as chemical degradation or in situ polymerisation were often observed. However, there may yet be some scope for further optimisation of the reaction conditions.

Michael addition also works well for the functionalisation of PAMA-based diblock copolymers. Appropriate derivatisation of selected PAMA-based block copolymers also allowed their GPC analysis in convenient eluents, whereas the precursors were sometimes not amenable to this technique. In addition, this facile chemistry has opened up routes to a wide range of functional block copolymers.

Acknowledgements

ESR thanks the University of Sheffield for an EPSRC PhD studentship. SPA is a recipient of a five-year Royal Society-Wolfson Research Merit Award. Unilever Research (Colworth) is thanked for CASE support of ESR and for permission to publish this work.

References

1. J. S. Wang and K. Matyjaszewski, J. Am. Chem. Soc., 1995, 117, 5614.
2. M. Kato, M. Kamigaito, M. Sawamoto and T. Higashimura, Macromolecules, 1995, 28, 1721.
3. M. Kamigaito, T. Ando and M. Sawamoto, Chem. Rev., 2001, 101, 3689.
4. K. Matyjaszewski and J. H. Xia, Chem. Rev., 2001, 101, 2921.
5. J. Chiefari, Y. K. Chong, F. Ercole, J. Krstina, J. Jeffery, T. P. T. Le, R. T. A. Mayadunne, G. F. Meijs, C. L. Moad, G. Moad, E. Rizzardo and S. H. Thang, Macromolecules, 1998, 31, 5559.
6. G. Moad, E. Rizzardo and S. H. Thang, Aust. J. Chem., 2005, 58, 379.
7. M. S. Donovan, A. B. Lowe, B. S. Sumerlin and C. L. McCormick, Macromolecules, 2002, 35, 4123.
8. C. L. McCormack and A. B. Lowe, Acc. Chem. Res., 2004, 37, 312.
9. L. Albertin, M. H. Stenzel, C. Barner-Kowollik, L. J. R. Foster and T. P. Davis, Macromolecules, 2005, 38, 9075.
10. I. Y. Ma, E. J. Lobb, N. C. Billingham, S. P. Armes, A. L. Lewis, A. W. Lloyd and J. Salvage, Macromolecules, 2002, 35, 9306.
11. M. Save, J. V. M. Weaver, S. P. Armes and P. McKenna, Macromolecules, 2002, 35, 1152.
12. J. V. M. Weaver, I. Bannister, K. L. Robinson, X. Bories-Azeau, S. P. Armes, M. Smallridge and P. McKenna, Macromolecules, 2004, 37, 2395.
13. Y. T. Li, S. P. Armes, X. P. Jin and S. P. Zhu, Macromolecules, 2003, 36, 8268.
14. X. Zhang, J. H. Xia and K. Matyjaszewski, Macromolecules, 1998, 31, 5167.
15. K. L. Beers, S. Boo, S. G. Gaynor and K. Matyjaszewski, Macromolecules, 1999, 32, 5772.
16. N. V. Tsarevsky, W. A. Braunecker, S. J. Brooks and K. Matyjaszewski, Macromolecules, 2006, 39, 6817.
17. M. H. Dufresne and J. C. Leroux, Pharm. Res., 2004, 21, 160.
18. K. L. Thompson, E. S. Read and S. P. Armes, Polym. Degrad. Stab., 2008, 93, 1460.
19. Y. T. Li, B. S. Lokitz and C. L. McCormick, Macromolecules, 2006, 39, 81.
20. L. H. He, E. S. Read, S. P. Armes and D. J. Adams, Macromolecules, 2007, 40, 4429.
21. R. Narain and S. P. Armes, Macromolecules, 2003, 36, 4675.
22. Y. T. Li and S. P. Armes, Macromolecules, 2009, 42, 939.
23. J. Z. Du and S. P. Armes, Langmuir, 2008, 24, 13710.
24. A. H. Alidedeoglu, A. W. York, C. L. McCormick and S. E. Morgan, J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 5405.
25. X. W. Xu, A. E. Smith, S. E. Kirkland and C. L. McCormick, Macromolecules, 2008, 41, 8429.
26. R. Narain and S. P. Armes, Biomacromolecules, 2003, 4, 1746.
27. J. Ramos, A. Martin-Molina, M. P. Sanz-Izquierdo, A. Rus, L. Borque, R. Hidalgo-Alvarez, F. Galisteo-Gonzalez and J. Forcada, J. Polym. Sci., Part A: Polym. Chem., 2003, 41, 2404.
28. G. Sharma and M. Ballauff, Macromol. Rapid Commun., 2004, 25, 547.
29. D. M. Haddleton, E. Depaquis, E. J. Kelly, D. Kukulj, S. R. Morsley, S. A. F. Bon, M. D. Eason and A. G. Steward, J. Polym. Sci., Part A: Polym. Chem., 2001, 39, 2378.
30. B. D. Mather, K. Viswanathan, K. M. Miller and T. E. Long, Prog. Polym. Sci., 2006, 31, 487.
31. P. Ferruti, M. A. Marchisio and R. Barbucci, Polymer, 1985, 26, 1336.
32. P. Ferruti, E. Ranucci, F. Trotta, E. Gianasi, E. G. Evagorou, M. Wasil, G. Wilson and R. Duncan, Macromol. Chem. Phys., 1999, 200, 1644.
33. P. Ferruti, M. A. Marchisio and R. Duncan, Macromol. Rapid Commun., 2002, 23, 332.
34. D. M. Lynn and R. Langer, J. Am. Chem. Soc., 2000, 122, 10761.
35. C. S. Wu, Y. L. Liu and Y. S. Chiu, Polymer, 2002, 43, 1773.
36. M. Morpurgo, F. M. Veronese, D. Kachensky and J. M. Harris, Bioconjugate Chem., 1996, 7, 363.
37. V. S. Khire, T. Y. Lee and C. N. Bowman, Macromolecules, 2007, 40, 5669.
38. D. L. Elbert, A. B. Pratt, M. P. Lutolf, S. Halstenberg and J. A. Hubbell, J. Controlled Release, 2001, 76, 11.
39. S. C. Rizzi and J. A. Hubbell, Biomacromolecules, 2005, 6, 1226.
40. S. A. Robb, B. H. Lee, R. McLemore and B. L. Vernon, Biomacromolecules, 2007, 8, 2294.
41. R. Roy and C. A. Laferriere, Chem. Commun., 1990, 1709.
42. D. L. Elbert and J. A. Hubbell, Biomacromolecules, 2001, 2, 430.
43. R. Duncan, S. Gac-Breton, R. Keane, R. Musila, Y. N. Sat, R. Satchi and F. Searle, J. Controlled Release, 2001, 74, 135.
44. S. Y. Liu and S. P. Armes, J. Am. Chem. Soc., 2001, 123, 9910.
45. K. Jankova, X. Y. Chen, J. Kops and W. Batsberg, Macromolecules, 1998, 31, 538.
46. In this context, it is interesting to note that He et al. found that well-defined PAMA–PDPA diblock copolymers could only be successfully synthesised via a ‘one-pot’ reaction provided that AMA was polymerised first.²⁰ In this earlier study, PDPA–PAMA could only be synthesised if PDPA homopolymer was first isolated and purified (i.e. the so-called macro-RAFT CTA approach) prior to chain extension with AMA, although this restriction may be mainly due to the differing solubility requirements of the two monomers and the corresponding polymers under the chosen RAFT conditions.
47. Y. Matsuda, M. Kobayashi, M. Annaka, K. Ishihara and A. Takahara, Polym. J., 2008, 40, 479.
48. Y. Kiritoshi and K. Ishihara, J. Biomater. Sci., Polym. Ed., 2002, 13, 213.
49. Y. Kiritoshi and K. Ishihara, Sci. Technol. Adv. Mater., 2003, 4, 93.
50. X. Bories-Azeau, S. P. Armes and H. J. W. van den Haak, Macromolecules, 2004, 37, 2348.
51. J. Clayden, N. Greeves, S. Warren and P. Wothers, Organic Chemistry, Oxford University Press, Oxford, UK, 2001.
52. E. S. Read and S. P. Armes, Chem. Commun., 2007, 3021.
53. J. Madsen, S. P. Armes, K. Bertal, H. Lomas, S. MacNeil and A. L. Lewis, Biomacromolecules, 2008, 9, 2265.

---