A clean synthesis approach to biocompatible amphiphilic conetworks via reversible addition–fragmentation chain transfer polymerization and thiol–ene chemistry

Abstract

A series of amphiphilic block copolymers containing hydrophobic polydimethylsiloxane (PDMS) segments and hydrophilic poly(N,N-dimethylacrylamide) (PDMAAm) segments have been synthesized by reversible addition–fragmentation chain transfer (RAFT) polymerization, which were then crosslinked into well-defined amphiphilic conetworks (APCNs) via ultraviolet (UV) induced thiol–ene click chemistry. Briefly, a PDMS-based RAFT agent was synthesized from the esterification of trithiocarbonate and bis(hydroxyethyloxypropyl) PDMS, and was used to control the RAFT polymerization of monomer DMAAm and allyl methacrylate (AMA) to form amphiphilic copolymers with a well-defined molecular mass and narrow dispersity. The amphiphilic copolymers were then crosslinked via UV induced thiol–ene click chemistry into APCNs, which showed unique amphiphilic characteristics as well as good mechanical properties, making them potential candidates in biomaterials. Transmission electron microscopy (TEM) and atomic force microscopy (AFM) inferred that the resultant APCN exhibited the behavior of microphase separation with a small channel size and uniform phase domain. Therefore, this kind of APCN possessed excellent comprehensive properties, i.e. a well-defined and co-continuous microstructure and high water uptake properties with a homogeneous hydrophilic channel, low cytotoxicity, high mechanical strength (2.1 ± 0.7 MPa) and elongation ratio (173 ± 17%), suggesting a promising biomaterial candidate for contact lenses, drug controlled systems, biomedical scaffolds for tissue engineering and supports for biocatalysts.

---

Introduction

Amphiphilic polymer conetworks (APCNs),^1–41 as a fairly important class of emerging polymeric materials, consisting of covalently interconnected hydrophilic and hydrophobic units with a continuous morphology have orthogonal properties, such as an amphiphilic structure, a nano-phase separated morphology, and semi-permeability,^3,4 and they can swell both in water and organic solvents and exhibit a large tendency for micro-phase separation due to thermally incompatible hydrophilic and hydrophobic segments.^5–10 The diverse functional behavior of APCNs makes them good candidates for various applications, including in contact lenses,¹¹ drug controlled release,^12,13 enzyme immobilization,¹⁴ separation membranes,^15,16 tissue engineering¹⁷ and other biomaterials.^18,19

Recently, polydimethylsiloxane (PDMS) has received great attention in APCN fabrication due to its unique characteristics such as low surface free energy, high elasticity, heat resistance, biological inertness and excellent biocompatibility as well as the highest oxygen permeability among other polymers,^20–22 and has been widely used with other units to form block and graft copolymers.^22–24 A wide variety of hydrophilic units, such as poly(ethylene glycol),^22,24 polyvinylpyrrolidone,²⁵ and (meth)acrylamides and functional (meth)acrylates^26–31 have been used. Poly(N,N-dimethylacrylamide) (PDMAAm), as one of the acrylamides, a commonly used hydrolytically stable, physiologically inert and biocompatible polymer, is deemed to be suitable for biomaterial applications.^32–34 Kennedy’s group prepared an APCN containing PDMAAm segments and studied its insulin permeability for semi-permeable artificial pancreas application,³⁵ which draws great attention to fabricating APCNs containing PDMS and PDMAAm.^32,36

Conventionally, most reported APCNs have been prepared via free radical polymerization between a macromonomer and monomer,^33,34,37,38 where macrophase separation will occur due to the thermodynamic incompatibility of the different segments and thermodynamic aggregation of the same segments, which inevitably leads to poor conformation regularity, structural defects or even loss of performance.^2,39 Although another optional method is to choose multifunctional polymers with a defined molecular mass,^{3,4,16,24,40,41} the random crosslinking between hydrophilic polymers and hydrophobic polymers often leads to some structural defects in the resulting APCNs. Besides, the limited types of polymers with multifunctional groups restrict its extensive application in APCN fabrication. To minimize these defects, many efforts have been made to seek effective approaches for the preparation of polymeric networks by cross-linking well-defined amphiphilic polymer chains via controlled methods.^9,10,36 Reversible addition–fragmentation chain transfer (RAFT) polymerization as one of the most powerful methods in the field of reversible deactivation radical polymerization,^42,43 has been widely used for the preparation of polymers with a well-defined molecular mass and narrow dispersity. In addition, RAFT polymerization has been proven to be a functional group tolerant method to control the polymerization process without any residual heavy metal ions, and has been extensively applied for the synthesis of various kinds of topologically structured copolymers with functional groups.^43–46 Meanwhile, the functional groups in well-defined copolymers have been explored due to their possible combination with click chemistry which would be a highly effective approach for the preparation of well-defined polymer co-networks due to their reaction specificity, high efficiency and functional group tolerance.^47–49 The newly emerging thiol–ene click chemistry induced by a thermal or photochemical process, without any residual heavy metal ions as encountered by the azide–alkyne click reaction, exhibits a more attractive application for biomaterial fabrication, where low toxicity and high efficiency are required especially in oral rehabilitation areas.^47–51 Thus, the combination of RAFT polymerization and thiol–ene click chemistry has been extensively used for the preparation of functional polymer materials with a well-defined molecular structure and high efficiency, and provides a convenient approach in the areas of new material fabrication.^52–54 Therefore, it is of great interest to fabricate an APCN containing PDMS/PDMAAm segments via RAFT polymerization which is crosslinked by thiol–ene click chemistry.

In this work, a straightforward method has been investigated to fabricate a PDMS based APCN by combining RAFT polymerization with thiol–ene click chemistry in a clean way with high efficiency. Briefly, a new PDMS-based macro-RAFT agent was synthesized to control the polymerization of N,N-dimethylacrylamide to obtain a triblock copolymer, which further controlled the polymerization of allyl methacrylate and was then modified to remove the bio-toxic thiocarbonylthio groups. The allyl groups in the resulting copolymer were crosslinked with pentaerythritol tetra(3-mercaptopropionate) to form the APCN via thiol–ene click chemistry induced by ultraviolet light. The resulting APCN with a well-defined molecular structure exhibits distinguished properties, i.e. unique swelling properties, low cytotoxicity, high mechanical properties as well as a microphase-separated morphology with a small channel size and uniform phase domain, suggesting a promising biomaterial in contact lenses, drug controlled systems, biomedical scaffolds for tissue engineering, supports for biocatalysts and especially semi-permeability materials for bioartificial pancreas fabrication.

Experimental

Materials

Azobisisobutyronitrile (AIBN) was recrystallized from methanol before use. N,N-Dimethylacrylamide (DMAAm) and allyl methacrylate (AMA) were purified by being passed over a column of basic alumina to remove any inhibitor. Tris(2-carboxyethyl)phosphine (TCEP, 98%), phosphate buffer solution (PBS, pH = 7.4), tert-butyl acrylate (t-BA, 99%) and n-hexamine were supplied by Aladdin Industrial Inc. without further purification. N,N-Dicyclohexylcarbodiimide (DCC, 99%), 4-dimethylaminopyridine (DMAP, 99%), and 2,2-dimethoxy-2-phenylacetophenone (DMPA, 98%) were purchased from J&K Scientific Ltd. Bis(hydroxyethyloxypropyl) polydimethylsiloxane (HO-PDMS-OH, DMS-C21, M_n = 4000 g mol⁻¹, D = 1.12) was purchased from Gelest. Pentaerythritol tetra(3-mercaptopropionate) (PETMP, 90%) was purchased from Tokyo Chemical Industry Co., Ltd. BHK-21 cells were purchased from the Shanghai Institute of Biochemistry Cell Biology. Anhydrous dichloromethane and tetrahydrofuran were distilled over calcium hydride before use. All reagents unless otherwise stated were purchased from Sinopharm Chemical Reagent Co., Ltd. and used as received without further purification.

Characterization

Gel permeation chromatography. GPC was performed on a BI-MwA gel permeation chromatograph (Waters, USA), equipped with a light scattering instrument (Brookhaven, USA) at room temperature, using THF as the eluent at a flow rate of 1 mL min⁻¹ with a polystyrene standard as the reference.

Proton nuclear magnetic resonance. ¹H NMR spectroscopy was performed on a Bruker Avance 400 instrument with CDCl₃ containing 1% TMS as the internal reference.

Ultraviolet irradiation. An Intelli-Ray 400 (Uvitron, USA) was used to induce the thiol–ene reaction under a light intensity of 50–100% for 0.5 h (λ = 365 nm).

Atomic force microscopy. AFM (E-SWEEP, Seiko, Japan) was performed for the imaging of the micro-phase separation of the APCN surfaces in the tapping mode. The surfaces of the dried samples were microtomed at room temperature with a diamond knife from Diatome and a Microtome ULTRACUTUCT (Leica), removing about 300 nm from the surface.

Transmission electron microscopy. TEM was performed on a JEOL JEM-2010 high-resolution transmission electron microscope at an acceleration voltage of 120 kV. Sections of the samples with 200–300 nm were trimmed using an ultrathin microtome machine before testing.

Scanning electron microscopy. SEM was performed on a JEOL JSM5600 scanning electron microscope using an accelerating voltage of 15 kV. The samples were then mounted on aluminium specimen stubs and sputter coated with gold before being examined.

Differential scan calorimetry. DSC thermographs were obtained using a DSC 204 F1 (Phoenix, Germany) instrument. The samples were quickly heated to 150 °C to eliminate their heat history and then quickly cooled to −150 °C with liquid nitrogen. After the preparation process, the DCS curves were recorded during the reheating process from −150 °C to 150 °C at a heating rate of 10 °C min⁻¹ under a nitrogen flow.

Tensile strength and elongation at break. Tensile strength and elongation at break were measured using a Universal Testing Machine (KEXIN, WDW3020, P. R. China). The samples were made into a rectangle (6 × 2 cm²) and the tensile speed was set at 10 mm min⁻¹. Each sample was measured three times and the average value was obtained. The error was less than 5%.

Cell culture. BHK-21 cells were cultured in a culture flask in RPMI-1640 medium filled with 1% L-glutamine, 1% penicillin/streptomycin and 10% fetal bovine serum, which then grew at 37 °C with 5% CO₂ for 3 days. Prior to cell culture, all samples (15 × 15 × 2 mm³) were placed in a 12-well plate and washed with PBS.

Synthesis of PDMS-based macro-RAFT

The chain transfer agent (CTA) S-1-dodecyl-S′-(α,α′-dimethyl-α′′-dimethyl-α′′-acetic acid) trithiocarbonate was synthesized according to the literature⁴⁵ with some modification. The PDMS based macro-RAFT was prepared by the esterification of CTA and HO-PDMS-OH in the presence of DCC/DMAP.⁵⁵ Procedures were performed as follows: CTA (2.2 g, 6.0 mmol) and DMAP (0.18 g, 1.5 mmol) were dissolved in 30 mL of anhydrous methylene chloride, then the mixture was stirred for 5 min before 4.0 g of HO-PDMS-OH (1 mmol) and 0.45 g of DCC (2.2 mmol) were added. The solution was stirred at room temperature for over 24 h, filtered through a Buchner funnel and passed through a silica gel column. The solution was concentrated using a rotary evaporator, re-filtered and washed several times to remove 1,3-dicyclohexylurea (DCU) and the unreacted CTA, and then dried under vacuum overnight to yield a transparent yellow oil. Yield: 3.91 g (63%). ¹H NMR (400 MHz, CDCl₃): δ 4.27 (t, 4H, CH₂CH₂OC=O), 3.64 (t, 4H, CH₂CH₂O), 3.43 (m, 4H, OCH₂CH₂), 3.29 (t, 4H, SCH₂(CH₂)₈), 1.73 (s, 12H, C(CH₃)₂), 1.57 (m, SiCH₂CH₂), 1.28 (m, CH₂(CH₂)₈CH₃), 0.9 (t, 6H, CH₃CH₂), 0.55 (m, 4H, SiCH₂CH₂), 0.1 (s, 6H, Si(CH₃)₂). M_n,GPC = 6800 g mol⁻¹, M_n,NMR = 4700 g mol⁻¹, D = 1.18.

Synthesis of triblock copolymer polydimethylacrylamide–polydimethylsiloxane–polydimethylacrylamide (PDMAAm–PDMS–PDMAAm)

A mixture of the macro-RAFT (1.46 g, 0.31 mmol), AIBN (6 mg, 0.037 mmol), and DMAAm (3.22 g, 25 mmol) in THF (20 mL) was placed in a round-bottom flask, deoxygenated by bubbling nitrogen for 30 min at room temperature and then immersed in a preheated oil bath. The polymerization proceeded at 65 °C for 6 h under constant magnetic stirring. The reaction mixture was quenched after being exposed to air. The polymer was precipitated into n-hexane three times to get a bright yellow solid, which was then moved into a vacuum oven and dried at 65 °C overnight. Yield: 3.83 g (82%). ¹H NMR (400 MHz, CDCl₃): δ 3.64 (t, 4H, CH₂CH₂O), 3.43 (m, OCH₂CH₂), 3.16 (t, 2H, CH₂CH(CON)S), 3.08 (s, 1H, SCH₂CH₂), 3.05–2.80 (m, 6H, N(CH₃)₂), 2.66 (s, 2H, CH₂CH(CON)S), 1.73 (s, 6H, C(CH₃)₂), 1.62 (m, 2H, SCH₂CH₂CH₂), 1.28 (m, CH₂(CH₂)₈CH₃), 0.92–0.86 (t, 5H, CH₃CH₂), 0.1 (s, 6H, Si(CH₃)₂). M_n,GPC = 2.88 × 10⁴ g mol⁻¹, M_n,NMR = 3.08 × 10⁴ g mol⁻¹, D = 1.30.

Synthesis of pentablock copolymer polyallylmethacrylate–polydimethylacrylamide–polydimethylsiloxane–polydimethylacrylamide–polyallylmethacrylate (PAMA–PDMAAm–PDMS–PDMAAm–PAMA)

A mixture of the triblock copolymer PDMAAm–PDMS–PDMAAm (6.16 g, 0.2 mmol), AMA (0.256 g, 2 mmol), AIBN (4 mg, 0.022 mmol), and THF (20 mL) was placed in a 100 mL round-bottom flask, deoxygenated by bubbling nitrogen for 30 min at room temperature and then immersed in a preheated oil bath. The polymerization proceeded at 65 °C for 11 h under constant magnetic stirring. The reaction mixture was quenched after being exposed to air. The polymer was precipitated into n-hexane three times to obtain a light yellow solid, which was dried under vacuum at 65 °C overnight. Yield: 5.61 g (87%). ¹H NMR (400 MHz, CDCl₃): δ 5.98 (s, 1H, CH₂CH=CH₂), 5.34 (d, 2H, CH=CH₂) 4.49 (s, 2H, CH₂CH=CH₂), 4.14 (m, CH₂CH₂OC=O), 3.16 (t, 2H, CH₂CH(CO)S), 3.08 (s, 1H, SCH₂CH₂), 3.05–2.80 (m, 6H, N(CH₃)₂), 2.66 (s, 2H, CH₂CH(CON)S), 1.80 (s, 3H, CH₂C(COO)(CH₃)S), 1.73 (s, 6H, C(CH₃)₂), 1.62 (m, 2H, SCH₂CH₂CH₂), 1.28 (m, CH₂(CH₂)₈CH₃), 1.07 (s, 3H, CH₂CH(CH₃)), 0.92–0.86 (t, 6H, CH₃CH₂), 0.49–0.58 (m, SiCH₂CH₂), 0.1 (s, 6H, Si(CH₃)₂). M_n,GPC = 3.16 × 10⁴ g mol⁻¹, M_n,NMR = 3.19 × 10⁴ g mol⁻¹, D = 1.33.

Post modification of PAMA–PDMAAm–PDMS–PDMAAm–PAMA

The pentablock copolymer PAMA–PDMAAm–PDMS–PDMAAm–PAMA (6.90 g, 0.2 mmol), n-butylamine (0.07 g, 1 mmol), and a trace amount of the reducing agent, TCEP, were dissolved in THF (20 mL). The solution was stirred for 1 h at room temperature under a nitrogen atmosphere, and changed from originally being yellow to colorless. An excess of tert-butyl acrylate (0.256 g, 2 mmol) was added and the reaction was left to proceed at room temperature for another 10 h. Afterwards, the solution was precipitated into n-hexane 3 times to get a colorless solid, which was then moved into a vacuum oven and dried at 65 °C overnight. Yield: 6.58 g (91%). ¹H NMR (400 MHz, CDCl₃): δ 5.98 (s, 1H, CH₂CH=CH₂), 5.34 (d, 2H, CH=CH₂) 4.49 (s, 2H, CH₂CH=CH₂), 4.14 (m, CH₂CH₂OC=O), 3.16 (t, 2H, SCH₂CH₂), 3.05–2.8 (m, 6H, N(CH₃)₂), 2.66 (s, 2H, CH₂CH(CON)S), 1.80 (s, 3H, CH₂C(COO)(CH₃)S), 1.73 (s, 6H, C(CH₃)₂), 1.37 (s, 9H, COO(CH₃)₃), 1.28 (m, CH₂(CH₂)₈CH₃), 1.07 (s, 3H, CH₂CH(CH₃)), 0.49–0.58 (m, SiCH₂CH₂), 0.1 (s, 6H, Si (CH₃)₂). M_n,GPC = 3.08 × 10⁴ g mol⁻¹, M_n,NMR = 3.40 × 10⁴ g mol⁻¹, D = 1.35.

Preparation of APCNs via thiol–ene click chemistry crosslinking

Post modified PAMA–PDMAAm–PDMS–PDMAAm–PAMA (1.0 g, 0.1 mmol), pentaerythritol tetra(3-mercaptopropionate) (0.326 g, 0.67 mmol) and DMPA (6 mg, 0.022 mmol) were dissolved in 2 mL of dichloromethane. The solution was moved into an Intelli-Ray 400 machine and kept under 90% light intensity at the light wavelength of 365 nm for 10 min. A colorless APCN membrane was obtained after the solvent evaporated.

Swelling measurement

The soluble content (Sol) in the APCN was recorded at room temperature when the weight of the APCN membranes remained unchanged after extracting with CHCl₃ several times. The following equation was used to express the data:^56,57where m_ex is the mass of the dry APCN membranes after extraction with CHCl₃ and m_dry is the original mass of the dry APCN membranes.

Swelling measurements were performed at room temperature by immersing the preweighed samples in excessive distilled water (or hexane). The swelling extent was measured by periodically moving samples from water (or hexane), removing the water (or hexane) absorbed to the surface by blotting with tissue paper and weighing. When the weight of the swollen samples remained unchanged for 40 h, the equilibrium swelling of the APCNs (S_w) was recorded. The following equation was used:

where m_sw is the mass of the swollen APCN membranes and m_dry is the mass of the dry APCN membranes after extraction.

Results and discussion

The synthesis strategy for the target APCN is shown in Fig. 1. Commercially available bis(hydroxyethyloxypropyl)polydimethylsiloxane is end-functionalized with S-1-dodecyl-S′-(α,α′-dimethyl-α′′-dimethyl-α′′-acetic acid)trithiocarbonate to form a PDMS-based macro-RAFT agent, which initiates the RAFT polymerization of N,N-dimethylacrylamide. The resulting triblock copolymer polydimethylacrylamide–polydimethylsiloxane–polydimethylacrylamide (PDMAAm–PDMS–PDMAAm) induces the RAFT polymerization of several AMA units to form the pentablock copolymer polyallylmethacrylate–polydimethylacrylamide–polydimethylsiloxane–polydimethylacrylamide–polyallylmethacrylate (PAMA–PDMAAm–PDMS–PDMAAm–PAMA). The bio-toxic trithiocarbonate groups in the resulting pentablock copolymer have been removed by reductive elimination in the presence of excessive n-butylamine, while the existing pendant AMA units provide the crosslinking points, where the terminal allyl groups can further react with the tetra-mercapto compounds to ensure the full crosslinking of the resulting APCN. RAFT polymerization guarantees the well-defined structure of the copolymers, which provides the prerequisite of a well-defined APCN. Therefore, the difference between the present work and the conetworks synthesized by the free radical copolymerization of functional macromolecules is the narrow molecular mass distribution and specific crosslinking points in comparison with the broad distribution and random crosslinking points in the conetworks synthesized by conventional free radical copolymerization.

Fig. 1 Synthesis strategy of the APCN.

In contrast to the previous work by Kennedy, who crosslinked an allyl-telechelic amphiphilic pentablock copolymer (PAMA-b-PDMAAm-b-PDMS-b-PDMAAm-b-PAMA) via hydrosilation with pentamethylcyclopentasiloxane (D₅H) in the presence of Karstedt’s catalyst, we adopted the UV light induced thiol–ene click reaction between the pentablock copolymer and pentaerythritol tetra(3-mercaptopropionate) under mild conditions, which also proves to be a rapid and efficient crosslinking method.⁵⁸

Polymer synthesis

The PDMS-based macro-RAFT agent was synthesized by coupling the commercially available precursor HO-PDMS-OH with trithiocarbonate CTA in the presence of DCC/DMAP. The ¹H NMR spectrum of the macro-RAFT agent in Fig. 2(B) shows a new chemical shift at 4.26 ppm assigned to –CH₂–O–C=O– which is newly formed during the esterification reaction and the absence of resonance at 3.74 ppm associated with the –CH₂–OH group in HO-PDMS-OH, indicating the successful conversion of HO-PDMS-OH to the macro-RAFT agent. Besides, the remaining shift at 0.90 ppm assigned to the trithiocarbonate groups suggests no loss of the trithiocarbonate groups after the esterification reaction. The obtained macro-RAFT agent is a bright yellow oil post purification, indicating that the trithiocarbonate groups were successfully transferred to the end of the PDMS chain.

Fig. 2 ¹H NMR spectra of intermediates: (A) HO-PDMS-OH, (B) PDMS-based macro-RAFT, (C) triblock copolymer PDMAAm–PDMS–PDMAAm, (D) pentablock copolymer PAMA–PDMAAm–PDMS–PDMAAm–PAMA and (E) modified PAMA–PDMAAm–PDMS–PDMAAm–PAMA.

The ¹H NMR spectrum in Fig. 2(C) gives further proof of the successful synthesis of the triblock copolymer PDMAAm–PDMS–PDMAAm, where the resonances at δ 3.2–2.8, 3.16 and 2.66 ppm are assigned to the methyl protons of the amide function –N(CH₃)₂, copolymer backbone methane protons and methylene protons, respectively. Additional resonances can be found at δ 0.1 and 0.49–0.58 ppm and are assigned to the methylene protons and siloxane methyl groups in the PDMS segments, respectively. The polymerizations of DMAAm in the presence of the PDMS-based macro-RAFT are performed at various temperatures to get optimal polymerization conditions for PDMAAm–PDMS–PDMAAm synthesis. As shown in Table 1, the molar mass and D of the copolymer go up with increasing temperature from 60 °C to 75 °C due to the increase of propagation radical concentration, leading to a faster polymerization rate and higher probability of bimolecular termination. The temperature is eventually set at 65 °C to obtain a copolymer with a narrow molecular mass distribution (D < 1.3) as well as a high average molar mass (M_n ∼ 30 kDa). The typical GPC trace of the copolymer is symmetrically monomodal (Fig. 3) and D is less than 1.5 (Table 1), indicating that the polymerizations of DMA proceed in a controlled manner. As RAFT polymerization occurs at a rate balance between activation (K_act) and transfer (K_trans), the rate of polymerization is not as high as conventional radical polymerization, which intensively relies on the transfer activity of the chain transfer agents. When the conversion reaches a high level, the rate of the transfer reaction slows down and the rate of any side reaction increases, leading to a higher D of the product. The M_n and D of the triblock copolymers goes up accordingly with the prolonged reaction time as more DMA units coupling into the polymer chains lower the chain transfer activity, increasing prolonged chain termination and broadening the molecular mass distribution.

Table 1 Effect of different temperatures and times on RAFT polymerization: [M]₀/[macro-RAFT]/AIBN = 320 : 1 : 0.2

No.	Temperature (°C)	Reaction time (h)	Mn,NMR^a × 10^−4	Mn,GPC^b × 10^−4	D	Polymer structure^c
a Calculated from the ^1H NMR based PDMS unit (6H at 0.1 ppm) and DMAAm unit (6H at 3.0 ppm).b Determined by GPC measurement using polystyrene standards.c For triblock copolymer PDMAAmx–PDMSy–PDMAAmx, the subscript means the number of the units, calculated from the ^1H NMR spectra.
1	60	8	1.25	1.06	1.19	PDMAAm40–PDMS54–PDMAAm40
2	65	8	3.08	2.88	1.30	PDMAAm132–PDMS54–PDMAAm132
3	70	8	3.52	3.06	1.39	PDMAAm154–PDMS54–PDMAAm154
4	75	8	3.67	3.29	1.50	PDMAAm163–PDMS54–PDMAAm163
5	65	4	1.75	2.18	1.17	PDMAAm65–PDMS54–PDMAAm65
6	65	6	2.52	2.32	1.25	PDMAAm103–PDMS54–PDMAAm103
7	65	8	3.08	2.88	1.30	PDMAAm132–PDMS54–PDMAAm132
8	65	10	3.35	3.09	1.38	PDMAAm145–PDMS54–PDMAAm145

---

Fig. 3 GPC traces of block polymers: (a) PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂ (D = 1.30), (b) PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄ (D = 1.33), (c) modified PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄ (D = 1.35) and (d) PDMS-based macro-RAFT agent (D = 1.18).

Allyl methacrylate (AMA), an asymmetrical divinyl monomer, has been copolymerized with a triblock copolymer via RAFT polymerization to obtain the amphiphilic pentablock copolymer with multiple reactive pendant allyl groups along the copolymer, which are readily crosslinked by thiol–ene click chemistry.

The successful synthesis of the key pentablock copolymer PAMA–PDMAAm–PDMS–PDMAAm–PAMA has been identified using ¹H NMR spectroscopy (Fig. 2(D)), where the resonances at δ 5.98, 5.34, and 4.49 ppm originate from the pendant allyl ester group (CH₂=CH–CH₂–O–C=O–). The spectrum also shows signals of the DMAAm units and PDMS segment at 2.96 and 0.1 ppm, which are ascribed to the proton resonance of the –N(CH₃)₂ groups and –Si(CH₃)₂–O– groups, respectively.

Moreover, the GPC traces of the intermediates have been characterized and the symmetrical monomodal GPC traces have been obtained as shown in Fig. 3. The narrow peak of the PDMS-based macro-RAFT agent indicates its narrow molecular mass dispersity (D = 1.18) as shown in Fig. 3(d). The molecular mass distribution of PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄ (D = 1.33) also shows a narrow peak almost the same as that of PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂ (D = 1.30) in the GPC traces as shown Fig. 3(a) and (b), respectively, indicating the polymerization of both DMAAm and AMA in a controlled manner.

Apart from the GPC traces, DSC traces have also been employed to confirm the block structure of the resulting copolymers. The triblock and pentablock copolymers were treated at the heating rate of 10 °C min⁻¹ from −150 °C to 150 °C, after quickly cooling down from 150 °C to eliminate the effect of the sample heat history during characterization, then the samples were reheated and the heating curves were collected to identify the glass transition temperature (T_g). As shown in Fig. 4, both the triblock copolymer PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂ and pentablock copolymer PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄ show two T_g, indicating the existence of two phases in the triblock and pentablock polymer chains, respectively. The glass transition temperature of the PDMAAm segment in the triblock copolymer PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂ is around 102.10 °C, which is less than that of the DMAAm homopolymer (113.00 °C) due to the plasticization effect of the PDMS segment (−118.26 °C).⁵⁹ Compared with the triblock copolymer PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂, the T_g of PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄ is shifted from −118.26 °C to −112.35 °C for the PDMS segment and from 102.10 °C to 108.50 °C for the PDMAAm segment, since the introduction of several AMA units leads to the thermal de-reactive movement by the slight elongation of the PAMA segment. Additionally, the exothermic peaks at around −80 °C and −90 °C in the DSC curves are due to the cold crystallization effects originating from the PDMS chains moving from the disordered state to the ordered state, resulting in energy release and the exhibition of exothermic peaks in the DSC curves.

Fig. 4 DSC traces of PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂ and PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄ pentablock copolymers.

As the thiocarbonylthio group has been deemed to be bio-poisonous in the field of biomaterial applications,⁶⁰ the thiocarbonylthio end group in the pentablock copolymer PAMA–PDMAAm–PDMS–PDMAAm–PAMA has been removed by reductive elimination with n-butylamine. This successful modification has been confirmed by the ¹H NMR spectrum in Fig. 2(E), which shows resonances at 4.49, 5.34 and 5.98 ppm ascribed to the pendant allyl groups, while the resonance at 0.9 ppm associated with thiocarbonylthio group disappears. Besides, the absence of the yellow color of the final product indicates the successful removal of the thiocarbonylthio group. Therefore, the pentablock copolymer shows no resonances associated with the thiocarbonylthio groups after butylamine aminolysis, suggesting the successful modification of the pentablock copolymer.

The GPC traces of the modified pentablock copolymer in Fig. 3(c) show little D difference with the original pentablock copolymer, illustrating the successful conversion.

UV light induced crosslinking

The remaining pendant allyl groups were crosslinked with an appropriate amount of PETMP (mercapto groups 1.2-fold stoichiometric excess to allyl groups) by thiol–ene click chemistry under UV light irradiation. After solvent evaporation and being extracted with DMF at room temperature for 24 h, an optically clear APCN membrane was obtained. The low Sol content (<5%) indicates efficient crosslinking, which demonstrates that the UV-induced thiol–ene crosslinking goes through a thorough conversion (Fig. 5).

Fig. 5 Appearance of the optically clear APCN membrane (PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄) in the dry state.

Swelling

The swelling data is a good predictor to evaluate the permeability of the APCN since the nutrient diffusion and swelling ratio of the APCN (S_w) are proportional to the volume fraction of the hydrophilic domain in the conetworks.^61,62 As shown in Fig. 6, the overall water S_w of the APCN membranes goes up with increasing PDMAAm content due to the excellent hydrophilicity of the PDMAAm segments, which even goes up to 700% with the increasing PDMAAm unit. When the APCN membrane consisted of PDMAAm₆₅–PDMS₅₄–PDMAAm₆₅, the water swelling even reaches 150% after 40 h, while the corresponding hexane swelling reaches a plateau of 50%. This swelling behavior was also consistent with the previous work by Patrickios and Tiller.^26,27 Therefore, the swelling behavior of the APCN varies with its composition and swelling medium, which confirms the amphiphilicity and co-continuous characteristics of the resulting APCN membranes.

Fig. 6 Effect of the hydrophilic segment length on the degree of swelling of the conetworks in water (a) and hexane (b).

Mechanical properties

As the human body is a water environment, the excellent mechanical strength for swollen membranes will guarantee their application in the area of bio-supports and bio-filtration.³² In order to investigate the mechanical properties, a series of APCN membranes with different PDMAAm amounts (wt%) have been synthesized, and their tensile strength and elongation ratio have been tested. As shown in Fig. 7, with the increasing PDMAAm content from 50% to 65%, the corresponding tensile strength of the swollen APCN membranes decreased sharply from 2.80 MPa to 1.39 MPa and the corresponding elongation ratio slightly decreases from 190% to 156%.

Fig. 7 Mechanical properties of APCN membranes with different PDMAAm content. Membrane properties: thickness, 150 ± 10 μm; swollen state.

Considering the relatively fine mechanical properties among the other swollen APCN membranes reported by J. Kang and coworkers (1.30–1.80 MPa, 45–60% elongation, 30–60% hydrophilic PDMAAm weight percentage),⁶³ the prepared PCN is a good candidate material in biomaterial areas.

AFM morphology

The APCN is designed to be a semi-permeable membrane in bioartificial pancreas fabrication. As we know, the hydrophilic segments in the APCN provide channels,⁶³ where nutrients diffuse to reach the pancreas cell, while the waste can be sent out. On the other hand, the hydrophobic PDMS segments provide channels for gas exchange, where oxygen can be taken in while the waste gases can be taken out. Therefore, a uniform channel size is important to guarantee permeation of the nutrients and oxygen for each transplanting cell to grow well.

The micro-phase separation morphology of the APCN membranes has been characterized using atomic force microscopy (AFM). According to the report of Bryan B. Sauer,⁶⁴ phase topology is intensively correlated with the segment properties, which can be sorted as hydrophilic and hydrophobic. As shown in Fig. 8, there appear obvious dark areas and bright areas in the APCN membrane during the AFM phase-mode scanning, which are related to the hydrophobic PDMS segments and hydrophilic segments, respectively. The dark areas assigned to the PDMS segments (around 10 nm) are much smaller than the bright ones (around 60 nm) assigned to the PDMAAm segments. This phase separated phenomenon was in accordance with the previous reports by Tiller and Patrickios.^27,30,65,66 A reasonable explanation is that the prolonged hydrophilic segment length is six times longer than the hydrophobic segments in the designed polymer (PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂), which further confirms the well-defined triblock copolymer. The obviously phase separated morphology and well controlled hydrophilic segments indicate that the resulting APCN may provide good biocompatibility and uniform channels for substance exchange in bioartificial pancreas fabrications.

Fig. 8 AFM images of the APCN consisting of the PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄ copolymer. (Left) Height; (Right) phase. Scan rate ,1 nm s⁻¹; scan box, 500 × 500 nm².

TEM morphology

Apart from AFM, the uniform micro-structure of the APCN has also been characterized using transmission electron microscopy (TEM) as shown in Fig. 9. The dark areas assigned to the PDMS segments are surrounded by the light areas assigned to the PDMAAm segments, forming a uniform “sea--island” structure with a relatively uniform size of dark “islands”, which originate from the well-defined structure of the block copolymers. The significant difference between the AFM and TEM images is the much smaller light domain sizes in the TEM images than in the AFM images, which originates from the scanning methods.⁴⁷ Both AFM and TEM methods confirm the uniform structure in the present APCN, which provides uniform channels in semi-permeable membrane fabrication. Therefore the combination of RAFT polymerization with thiol–ene is an effective method to fabricate APCN membranes with an almost precisely controlled microstructure.

Fig. 9 TEM image of the swollen APCN consisting of the PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄ copolymer.

Cell culture

As this APCN has been designed to be applied in biomaterial areas, its lack of cell toxicity is necessary for its biomedical safety. As shown in Fig. 10, BHK-21 cells reproduce well on the APCN materials and the survival rate of cells on the APCN sample is over 95% after two weeks as compared with the control one, which infers its relatively low toxicity to the cultured cells. The good biocompatibility of these materials indicates the accessibility to fabricate biomedical materials using this method.

Fig. 10 (Left) Cell colony morphology after immersing APCN membranes in cell-culture medium (×1000, sample); (Right) cell colony morphology in cell-culture medium (×1000, control). APCN consisted of the PAMA₄–PDMAAm₁₃₂–PDMS₅₄–PDMAAm₁₃₂–PAMA₄ copolymer; cell colony morphologies were taken using SEM.

Conclusions

Novel amphiphilic conetworks containing PDMAAm and PDMS segments are fabricated through thiol–ene click end-crosslinking with well-defined allyl terminated amphiphilic pentablock copolymers via RAFT polymerizations. The resulting APCN exhibits unique amphiphilic characteristics, i.e. controllable swelling behavior, a microphase separated morphology and excellent mechanical properties. Moreover, a cell culture test indicates that the target APCN possesses low cytotoxicity and good biocompatibility. Therefore, this work provides an accessible method to fabricate a series of APCNs with controlled structure and low cytotoxicity by the combination of RAFT polymerization with non-Cu²⁺ click chemistry, and their diverse and excellent properties are suitable for potential biomaterial applications.

Acknowledgements

We thank the great financial supports of the National Science Foundation (NSF51103019, NSF21174027), National High-tech Research and Development Projects (863, 2012AA03A605) and Program for New Century Excellent Talents in University (12X10623).

Notes and references

1. G. Erdodi and J. P. Kennedy, Prog. Polym. Sci., 2006, 31, 1.
2. C. S. Patrickios and T. K. Georgiou, Curr. Opin. Colloid Interface Sci., 2003, 8, 76.
3. C. Fodor, G. Kali and B. Iván, Macromolecules, 2011, 44, 4496.
4. C. Fodor, J. Bozi, M. Blazsó and B. Iván, RSC Adv., 2015, 5, 17413.
5. J. Scherble, R. Thomann, B. Iván and R. Mülhaupt, J. Polym. Sci., Part B: Polym. Phys., 2001, 39, 1429.
6. B. Iván, K. Almdal, K. Mortensen, I. Johannsen and J. Kops, Macromolecules, 2001, 34, 1579.
7. A. Domján, G. Erdodi, M. Wilhelm, M. Neidhöfer, K. Landfester, B. Iván and H. W. Spiess, Macromolecules, 2003, 36, 9107.
8. B. Iván, M. Haraszti, G. Erdodi, J. Scherble, R. Thomann and R. Mulhaupt, Macromol. Symp., 2005, 227, 265.
9. N. Bruns, J. Scherble, L. Hartmann, R. Thomann, B. Iván, R. Mülhaupt and J. C. Tiller, Macromolecules, 2005, 38, 2431.
10. G. Kali, T. K. Georgiou, B. Iván and C. S. Patrickios, J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 4289.
11. J. Xu, X. Li and F. Sun, Drug Delivery, 2011, 18, 150.
12. N. Ghasdian, E. Church, A. P. Cottam, K. Hornsby, M. Y. Leung and T. K. Georgiou, RSC Adv., 2013, 3, 19070.
13. C. Lin and I. Gitsov, Macromolecules, 2010, 43, 10017.
14. N. Bruns and J. C. Tiller, Nano Lett., 2005, 5, 45.
15. J. Tobis, L. Boch, Y. Thomann and J. C. Tiller, J. Membr. Sci., 2011, 372, 219.
16. S. K. Jewrajka, G. Erdodi, J. P. Kennedy, D. Ely, G. Dunphy, S. Boehme and F. Popescu, J. Biomed. Mater. Res., Part A, 2008, 87, 69.
17. Y. Sun, J. Collett, N. J. Fullwood, S. M. Neil and S. Rimmer, Biomaterials, 2007, 28, 661.
18. S. Meskath, G. Urban and J. Heinze, Sens. Actuators, B, 2013, 186, 367.
19. Y. Wang, J. A. Finlay, D. E. Betts, T. J. Merkel, J. C. Luft, M. E. Callow, J. A. Callow and J. M. DeSimone, Langmuir, 2011, 27, 10365.
20. R. Wang, Y. L. Yang, M. Qin, L. K. Wang, L. Yu, B. Shao, M. Q. Qiao, C. Wang and X. Z. Feng, Chem. Mater., 2007, 19, 3227.
21. A. L. Cordeiro, S. Zschoche, A. Janke, M. Nitschke and C. Werner, Langmuir, 2009, 25, 1509.
22. G. Lin, X. Zhang, S. R. Kumar and J. E. Mark, Silicon, 2009, 1, 173.
23. E. Pouget, J. Tonnar, P. Lucas, P. Lacroix-Desmazes, F. Ganachaud and B. Boutevin, Chem. Rev., 2009, 110, 1233.
24. G. Erdodi and J. P. Kennedy, J. Polym. Sci., Part A: Polym. Chem., 2005, 43, 4953.
25. V. P. Torchilina, T. S. Levchenkoa, K. R. Whitemana, A. A. Yaroslavovb, A. M. Tsatsakisc, A. K. Rizosb, E. V. Michailovad and M. I. Shtilmand, Biomaterials, 2001, 22, 3035.
26. D. S. Achilleos, T. K. Georgiou and C. S. Patrickios, Biomacromolecules, 2006, 7, 3396.
27. G. Kali, T. K. Georgiou, B. Ivan, C. S. Patrickios, E. Loizou, Y. Thomann and J. C. Tiller, Langmuir, 2007, 23, 10746.
28. T. K. Georgiou, C. S. Patrickios, P. W. Groh and B. Ivan, Macromolecules, 2007, 40, 2335.
29. G. Kali, T. K. Georgiou, B. Iván and C. S. Patrickios, J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 4289.
30. G. Kali, T. K. Georgiou, B. Iván, C. S. Patrickios, E. Loizou, Y. Thomann and J. C. Tiller, Macromolecules, 2007, 40, 2192.
31. M. Achilleos, T. Krasia-Christoforou and C. S. Patrickios, Macromolecules, 2007, 40, 5575.
32. J. Kang, G. Erdodi, J. P. Kennedy, H. Chou, L. Lu and S. Grundfest-Broniatowski, Macromol. Biosci., 2010, 10, 369.
33. B. Ivan, J. P. Kennedy and P. W. Mackey, ACS Symp. Ser., 1991, 469, 194.
34. B. Ivan, J. P. Kennedy and P. W. Mackey, Amphiphilic Networks, US Pat., 5,073,381, Dec. 17, 1991.
35. J. P. Kennedy, G. Fenyvesi, S. Na, B. Keszler and K. S. Rosenthal, Des. Monomers Polym., 2000, 3, 113.
36. R. Karunakaran and J. P. Kennedy, J. Polym. Sci., Part A: Polym. Chem., 2007, 45, 4284.
37. J. Scherble, B. Ivan and R. Mulhaupt, Macromol. Chem. Phys., 2002, 203, 1866.
38. M. Haraszti, E. Tóth and B. Ivan, Chem. Mater., 2006, 18, 4952.
39. G. Kali, S. Vavra, K. László and B. Ivan, Macromolecules, 2013, 46, 5337.
40. G. Erdodi and B. Ivan, Chem. Mater., 2004, 16, 959.
41. W. H. Binder, L. Petraru, T. Roth, P. W. Groh, V. Pálfi, S. Keki and B. Ivan, Adv. Funct. Mater., 2007, 17, 1317.
42. A. D. Jenkins, R. G. Jones and G. Moad, Pure Appl. Chem., 2009, 82, 483.
43. G. Moad, E. Rizzardo and S. H. Thang, Aust. J. Chem., 2012, 65, 985.
44. 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. Tang, Macromolecules, 1998, 31, 5559.
45. J. T. Lai, D. Filla and R. Shea, Macromolecules, 2002, 35, 6754.
46. M. R. Hill, R. N. Carmean and B. S. Sumerlin, Macromolecules, 2015, 48, 5459.
47. A. B. Lowe, Polym. Chem., 2010, 1, 17.
48. S. P. S. Koo, M. M. Stamenović, R. A. Prasath, A. J. Inglis, F. E. Du Prez, C. Barner-Kowollik, W. V. Camp and T. Junkers, J. Polym. Sci., Part A: Polym. Chem., 2010, 48, 1699.
49. C. N. Walker, K. C. Bryson, R. C. Hayward and G. N. Tew, ACS Nano, 2014, 8, 12376.
50. C. E. Hoyle and C. N. Bowman, Angew. Chem., Int. Ed., 2010, 49, 1540.
51. L. M. Campos, K. L. Killops, R. Sakai, J. M. J. Paulusse, D. Damiron, E. Drockenmuller, B. W. Messmore and C. J. Hawker, Macromolecules, 2008, 41, 7063.
52. C. Boyer, A. Granville, T. P. Davis and V. Bulmus, J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 3773.
53. J. W. Chan, B. Yu, C. E. Hoyle and A. B. Lowe, Chem. Commun., 2008, 40, 4959.
54. C. Boyer and T. P. Davis, Chem. Commun., 2009, 40, 6029.
55. M. L. Wadley and K. A. Cavicchi, J. Appl. Polym. Sci., 2010, 115, 635.
56. J. P. Kennedy, K. S. Rosenthal and B. Kashibhatla, Des. Monomers Polym., 2004, 7, 485.
57. G. Erdodi and J. P. Kennedy, J. Polym. Sci., Part A: Polym. Chem., 2007, 45, 295.
58. R. Karunakaran and J. P. Kennedy, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 4254.
59. D. Pavlović, J. G. Linhardt, J. F. Künzler and D. A. Shipp, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 7033.
60. C. Boyer, V. Bulmus, T. P. Davis, V. Ladmiral, J. Liu and S. Perrier, Chem. Rev., 2009, 109, 5402.
61. T. Canal and N. A. Peppas, J. Biomed. Mater. Res., 1989, 23, 1183.
62. S. R. Lustig and N. A. Peppas, J. Appl. Polym. Sci., 1988, 36, 735.
63. J. Kang, G. Erdodi and J. P. Kennedy, J. Polym. Sci., Part A: Polym. Chem., 2007, 45, 4276.
64. B. B. Sauer, R. S. McLean and R. R. Thomas, Langmuir, 1998, 14, 3045.
65. N. Bruns and J. C. Tiller, Nano Lett., 2005, 5, 45.
66. J. Tobis, Y. Thomann and J. C. Tiller, Polymer, 2010, 51, 35.

---

Footnote

† These authors contributed equally.

---