Article pubs.acs.org/Macromolecules
Effect of Molecular Parameters on the Architecture and Membrane Properties of 3D Assemblies of Amphiphilic Copolymers Dalin Wu,† Mariana Spulber,† Fabian Itel,† Mohamed Chami,‡ Thomas Pfohl,† Cornelia G. Palivan,*,† and Wolfgang Meier*,† †
Department of Chemistry, University of Basel, Klingelbergstrasse 80, 4056 Basel, Switzerland Centre for Cellular Imaging and Nano Analytics, Biozentrum, University of Basel, Mattenstrasse 26, 4058 Basel, Switzerland
‡
S Supporting Information *
ABSTRACT: Reliable prediction of the 3D structure of self-assembled amphiphilic copolymers is essential for applications in which specificity has to be carefully controlled, as for example in nanomedicine. Since supramolecular assemblies are strongly affected by the chemical nature of block copolymers and the preparation methods, it is essential to understand the influence of such parameters on the self-assembly process. We have now successfully synthesized a library of amphiphilic block copolymers, poly(dimethylsiloxane)-block-poly(2-methyl-2-oxazoline) (PDMS-b-PMOXA), and investigated the molecular parameters and self-assembly conditions that generate a specific architecture in particular polymersome. We found that 3D assemblies are strongly affected by the preparation method, but not by the initial concentration of block copolymer and solution pH. The phase diagrams of selfassembly behavior show a strong influence of the hydrophilic to hydrophobic ratio ( f PMOXA), and the molecular mass of each block. In particular, the formation of polymersomes was possible only for block copolymers with high molecular mass PDMS and low f PMOXA values. A combination of very low molecular mass PDMS and small f PMOXA values induced formation of worm-like micelles, while low molecular mass PDMS and high f PMOXA values generated a mixture of small micelles and spherical particles. The polymersome membranes were characterized by electron paramagnetic resonance, which indicated a multilayer structure (PMOXA outer layer, PDMS middle part, and PMOXA inner layer) with low flexibility and permeability. Addition of detergent increased both the flexibility and permeability of the PDMS block, which was proven to act as a barrier layer for the membranes.
■
INTRODUCTION Independent self-organization of various structures into patterns and hierarchic structures with different architectures represents one of the driving forces of life in terms of specific structural and functional roles. Understanding and mimicking natural self-assembled structures based on low molecular interactions (van der Waals, ionic interactions), are now in focus, and are useful for the development of templates, membranes, and compartments in various domains, such as catalysis, medicine, electronics, and technology.1,2 In addition, synthetic supramolecular assemblies, if appropriately designed can be combined with biological molecules (proteins, DNA, enzymes, mimics) to create hybrid materials, which benefit from the stability of synthetic materials and the activity of the biological entities.3,4 In this respect, amphiphilic block copolymers represent one of the most promising selfassembling materials, because they generate in dilute aqueous solutions a variety of supramolecular assemblies, such as polymersomes, free-standing films, tubes, micelles, or hard spheres.5 The architectures and properties of these synthetic supramolecular assemblies can be changed by chemical modification of copolymer blocks (chemical nature, block lengths, or functionalization), and modulation of their hydrophilic-to-hydrophobic ratio.6,7 © 2014 American Chemical Society
An interesting architecture in terms of possible novel applications is based on polymer vesicles (polymersomes), which are nanometer size compartments generated by selfassembly in dilute aqueous conditions. They offer three different topological regions (membrane, external surface, and inner cavity) that serve as appropriate locations for active molecules, such as proteins, enzymes, mimics, DNA, contrast agents, etc.8−10 The hydrophobic membrane facilitates insertion of hydrophobic molecules, the external surface can be functionalized with specific molecules to support targeting approaches or immobilization on solid supports, while the inner cavity has the role of a confined space where hydrophilic active molecules can be encapsulated. A major advantage of polymersomes, which supports their use as drug delivery systems, is their dual role: they prevent leakage of encapsulated molecules and protect them against exterior degradation agents.11 A further step in the development of applications for polymersomes was to design nanoreactors by using the confined space, in which reactions occur when catalytically active compounds are encapsulated inside it.12 To support Received: March 10, 2014 Revised: June 30, 2014 Published: July 16, 2014 5060
dx.doi.org/10.1021/ma500511r | Macromolecules 2014, 47, 5060−5069
Macromolecules
Article
interaction of spin-probes with the polymer membranes of the polymersomes by electron paramagnetic resonance (EPR) to provide information on local flexibility and polarity of the polymer membranes. Such information is essential when planning the insertion of membrane proteins into polymersome membranes for nanoreactor development.
nanoreactor functionality, the polymersome membrane has to allow an exchange of substrates/products with the environment while preserving the spherical architecture of vesicles. Permeability of the membrane is achieved either by chemical modification,13,14 or by insertion of channel proteins.12 A large variety of nanoreactors have been produced by encapsulation of proteins,15 enzymes,16 mimics,17 or combinations of thereof that enable the development from conventional drug delivery systems, up to artificial organelles18 when nanoreactors mimic natural organelles inside cells.9 As supramolecular assemblies with a variety of architectures (micelles, polymersomes, worm-like structures, nanotubes) can be generated by the self-assembly process, an intimate understanding of factors that influence this process is essential in order to be able to generate desired structures. However, the self-assembly process is governed by a complex scenario involving various factors such as (i) the chemical composition of amphiphilic block copolymers, (ii) the methods of preparation, and (iii) external factors (pH, temperature, the presence of active compounds to be inserted/encapsulated). The combination of these factors induces the formation of a variety of supramolecular morphologies.19−22 The relationship between the self-assembled architectures and their chemical structures and external factors, requires focus on both theoretical and experimental research in order to be able to produce the desired architecture for an intended application.23 One of the factors that has been considered as determining the shape of self-assembled supramolecular structures is the ratio between the hydrophilic block and the hydrophobic block, which determines the curvature of the hydrophilic−hydrophobic interface.6 The influence of the hydrophilic/hydrophobic ratio was studied for poly(butadiene)-block-poly(ethylene oxide) (PBD-b-PEO) block copolymers, whose morphologies change from spherical micelles to wormlike micelles and finally to polymersomes on decreasing the hydrophilic/hydrophobic block ratio.24 The different structures of the various supramolecular assemblies formed by changing the hydrophilic/hydrophobic ratio has been proposed to be the consequence of the interfacial area increasing with hydrophilic block length, although no other factors were simultaneously taken into account.25 However, the influence of the factors, such as polydispersity index (PDI), pH of polymer solutions, the presence of ions, and the concentration of block copolymers has been investigated separately for various types of block copolymers.19,20,26,27 Relationships between the architecture of supramolecular assemblies of amphiphilic block copolymers and various parameters have been reported for several block copolymer systems;6,11,19,20,22,28−33 however, in depth and systematic analyses of various factors simultaneously affecting their self-assembly are still needed. In the present work, we have synthesized a library of amphiphilic block copolymers poly(dimethylsiloxane)-blockpoly(2-methyl-2-oxazoline) (PDMS-b-PMOXA), that are designed specifically to facilitate the study of the effect of the molecular mass of both hydrophobic and hydrophilic blocks on the self-assembly behavior. The initial concentrations of copolymers, preparation methods for the supramolecular assemblies, and the contents of the aqueous solution are taken into account, and a combination of light scattering, TEM and cryoTEM have been used to characterize the supramolecular assemblies. In addition, in order to obtain greater insight into the organization of the hydrophilic and hydrophobic domains of the polymersomes, we have investigated the
■
EXPERIMENTAL SECTION
Materials and Synthesis. Materials. 2-Methyl-2-oxazoline (98%, Aldrich), triethylamine (≥99%, Aldrich), trifluoromethanesulfonic anhydride (≥99%, Aldrich), hexamethylcyclotrisiloxane (D3) (98%, Aldrich), calcium hydroxide (95%, Aldrich), platinum(0)-1,3-divinyl1,1,3,3-tetramethyldisiloxane complex (Pt(dvs)) solution in xylene (Pt ∼2%, Aldrich), 2-allyloxyethanol (98%, Aldrich), chlorodimethylsilane (98%, Aldrich), n-butyllithium lithium solution (2.5 M in cyclohexane, Aldrich), activated carbon (Fluka), molecular sieve 4 Å (Chemie Uetikon AG, Switzerland), 5-doxylstearic acid (5-DSA), and 16doxylstearic acid (16-DSA). Monocarbinol poly(dimethylsiloxane) with molecular mass 5 kDa was purchased from ABCR. Hexamethylcyclotrisiloxane (D3), 2-methyl-2-oxazoline, and triethylamine were freshly distilled from calcium hydroxide under argon before use. Toluene, acetonitrile, and chloroform with 100−200 ppm amylenes as stabilizers were refluxed with calcium hydroxide under argon. Tetrahydrofuran (THF) and cyclohexane were refluxed with sodium and benzophenone under argon. Triton X-100 (Fluka) was used as a 1% solution in distilled water. Other chemicals and solvents were used directly without any purification. Synthesis of Monocarbinol Poly(dimethylsiloxane) (PDMS−OH). PDMS−OH was synthesized according to the procedure reported by Hideki Kazamaet al.34 using the platinum(0)-1,3-divinyl-1,1,3,3tetramethyldisiloxane complex (Pt(dvs)) as the catalyst during the hydrosilylation reaction (Supporting Information, Figure S1B). S y n t h e s i s o f PD MS - O T f Ma c r o i n i t i a t o r an d P o l y (dimethylsiloxane)-block-Poly(2-methyl-2-oxazoline) (PDMS-bPMOXA) Diblock Copolymers. PDMS-OTf macroinitiator and PDMS-b-PMOXA diblock copolymers were synthesized according to a modified procedure of the synthesis route reported by Egli et al. (Supporting Information, Scheme S1).35 Chloroform/acetonitrile (v/v = 7:3) was used as reaction medium for the polymerization of 2methyl-2-oxazoline, while H2O and triethylamine were used as quenching agents instead of KOH/MeOH solutions (Supporting Information, Figure S1C and Figure S2). Instruments and Measurements. 1H NMR spectra were recorded on a Bruker DPX-400 MHz spectrometer in CDCl3 without tetramethysilane standard and analyzed using MestReNova software. The molecular mass and polydispersity index of PDMS were determined using a Viscotek GPC max system equipped with four Agilent PL gel columns (10 μm guard; mixed C; 10 μm, 100 Å; 5 μm, 103 Å). THF was used as eluent at a flow rate of 1 mL min−1 at 40 °C. Signals were recorded with a refractive-index detector and calibrated against polystyrene standards (Agilent). Transmission electron microscopy (TEM) was used to analyze the supramolecular assemblies of block copolymers. Sample solutions were negatively stained with 2% uranyl acetate solution and deposited on a carbon-coated copper grid. The samples were examined on a Philips Morgagni 268D TEM operated at 80 kV. Particle sizes were calculated from TEM images by averaging the diameters (mean ± S.D.) from at least 50 particles. Cryo-electron microscopy was used to visualize the self-assembled structures in their native environment. Polymer suspensions in buffer (20 mM Hepes, pH 7.4, 50 mM NaCl) at concentrations of 5 mg/mL were deposited on glow-discharged holey carbon grids (Quantufoil, Germany) and blotted before quick-freezing in liquid ethane by using a Vitrobot plunging freezing device (FEI Company, USA). The grids were stored in liquid nitrogen before transferring them to a cryoholder (Gatan, USA).). Imaging was performed with a Philips CM200 FEG TEM at an accelerating voltage of 200 kV in low-dose mode with a defocus of −6 μm for imaging and a defocus of −3 μm for membrane thickness determination. Membrane thicknesses represent 5061
dx.doi.org/10.1021/ma500511r | Macromolecules 2014, 47, 5060−5069
Macromolecules
Article
a mean value of at least 150 single distance measurements along the membrane thickness of five different images.36 Light scattering (LS) was used to characterize the assembled structures in solution. Dynamic light scattering (DLS) and static light scattering (SLS) experiments were performed on an ALV goniometer (ALV GmbH, Germany), equipped with an ALV He−Ne laser (λ = 632.8 nm). Light scattering measurements were performed in 10 mm cylindrical quartz cells at angles of 30−150° at 293 K. The obtained data were processed using ALV static and dynamic fit and plot software (version 4.31 10/01). SLS data were processed according to the Guinier-model, and DLS data by using a Williams−Watts function. The critical micelle concentration (CMC) was determined for the PDMS65-b-PMOXA14 and PDMS65-b-PMOXA32 by surface tension with concentration from 3 × 10−4 to 1 × 10−2 mg/mL. Surface tension was measured on a Sigma 703D (KSV Inst.) tensiometer with a platinum wilhelmy plate precleaned with isopropyl alcohol and water, followed by flame annealing. Solution was prepared 24 h prior to measurement by diluting the stock solution to different concentration. The CMC values were calculated from the interception point of the two tangent straight lines. Small angle X-ray scattering (SAXS) was performed on a Bruker AXS Nanostar with an Incoatec Cu-IμS Microfocus X-ray source (λ = 0.154 nm) and a virtually noise-free, real-time 2D Hi-Star detector with photon counting ability. The measurements were done with 45 kV and 650 μA and the integration time was 9 h. About 20 μL of sample with the concentration of 5 mg·mL−1 was loaded into a glass capillary (d = 1 mm, thickness of the wall = 0.01 mm). The closed capillary set up into the Nanostar, vacuum was applied, and the measurement was started. The data were azimuthally averaged with SAXS v.4.1.36 Bruker software and fitted with Nanofit. Electron paramagnetic resonance (EPR) measurements were performed on a Bruker CW EPR Elexsys-500 spectrometer equipped with a variable temperature unit. The spectra were recorded at temperatures varying between 150 and 320 K with the following parameters: 100 kHz magnetic field modulation, microwave power 2 mW, number of scans up to 20, modulation amplitude in the range of 0.4 G. The nitrogen hyperfine coupling (aN) was determined directly from the spectra for motionally narrowed lineshapes with an error limit of 5%. 2aN values correspond to the distance (in gauss) between the low-field and high-field lines of the motionally averaged spectra (triplets) recorded at temperatures above 270 K, whereas 2Azz values are measured as a distance between the last minimum and the first maximum (extreme separation) in the rigid limit spectra recorded from frozen solutions. The microviscosity in the proximity of the nitroxide free-radical probe was determined from the correlation time τc, which is related to the rate of rotational reorientation of the probe
⎧ I τc = 6.5e−10 × ΔH0 × ⎨ 0 + ⎩ I+1 ⎪
⎪
mM sucrose solution, and closed with a second ITO glass plate. GUVs were generated at 25 °C with a frequency of 3.0 Hz and amplitude of 2.5 V for 3 h. For visualization, a few microliters of the GUV solution, stained with Bodipy 630/650, were dropped into a microscopy chamber filled with 200 μL of buffer (20 mM Hepes, pH 7.4, 50 mM NaCl) and investigated at 20 °C using a confocal laser scanning microscope (Zeiss LSM 510-META/Confocor2). Preparation of Polymersomes Containing Spin Probes. Polymersomes containing spin probes were prepared by two methods: film rehydration with 0.1 mM 5 DSA solution in 0.1 M NaOH, and direct mixture of the polymersome solution obtained by film rehydration of four different PDMS-b-PMOXA diblock copolymers, PDMS65-b-PMOXA12, PDMS65-b-PMOXA14, PDMS65-b-PMOXA19, and PDMS65-b-PMOXA32, with 5 DSA and PDMS65-b-PMOXA12, PDMS65-b-PMOXA14, and PDMS65-b-PMOXA32 with 16 DSA weighted so that the final concentration of spin probe was 1 mM. The spin probes were dissolved in ethanol to a concentration of 1 mM and divided into several vials. After evaporation of the solvent in a stream of nitrogen, the spin film was dissolved in a neat polymersome solution prepared in the previous step, and the solutions stirred overnight before being transferred to capillary tubes. The spin probe concentration in polymersome solutions was kept as 0.1 mM in order to prevent line width broadening by spin−spin interactions.39
■
RESULTS AND DISCUSSION
A library of PDMSx-b-PMOXAy block copolymers were synthesized for investigating the influence of their molecular properties on the architecture and properties of the supramolecular assemblies formed by self-assembly. We selected three hydrophobic PDMS polymers with increasing degree of polymerization: 16, 39, and 65 units and narrow polydispersity indexes: 1.11, 1.08, 1.12, respectively (Supporting Information, Figure S3). To each of them, hydrophilic blocks PMOXA of variable block lengths (ranging from 3 to 38 units) were coupled by cationic ring-opening polymerization of 2-methyl-2oxazoline. The PDMS-OTf macroinitiator, necessary to initiate the ringopening polymerization of 2-methyl-2-oxazoline, was obtained by treating PDMS−OH with trifluoromethane sulfonic anhydride and dry TEA (Supporting Information, Figure S1C). The chemical shift of the −CH2−OH peak from 3.72 to 4.62 ppm, indicated that no free hydroxyl groups remained after activation of PDMS−OH (Supporting Information, Figure S1A, B). The library of PDMSx-b-PMOXAy diblock copolymers allowed modulation of the formation of supramolecular assemblies obtained by self-assembly in dilute aqueous solution (Table 1) by varying the hydrophilic to hydrophobic block length ratio (fPMOXA). The values of f PMOXA are the molecular mass of PMOXA divided by the molecular mass of PDMS, and the values of f* are the molecular mass of PMOXA divided by the molecular mass of the whole block copolymer. An increase in fPMOXA and f * indicates an increase in the length of the hydrophilic PMOXA block. The film rehydration method was used to obtain supramolecular self-assemblies because it is known to promote the formation of polymersomes.40 Polymer self-assemblies were characterized by a combination of light scattering and transmission electron microscopy. We were interested to establish whether there is a relation between f PMOXA or the length of each polymer block and a specific architecture of the supramolecular assemblies. Effect of f PMOXA on the Architecture of Supramolecular Assemblies. Three types of phase diagrams were obtained based on the molecular masses of PDMS, and f PMOXA
⎫ I0 − 2⎬ I −1 ⎭ ⎪
⎪
where ΔH0 is the line width of the mI = 0 transition, and I0, I+1, I−1 are the peak to peak heights of the mI = 0, +1, and −1 transitions, respectively.37 Methods for Preparation of Supramolecular Assemblies. Block copolymers were self-assembled in aqueous solution by three procedures known as the film rehydration, cosolvent, and electroformation methods. The film rehydration method involved dissolution of the copolymer (5 mg) in EtOH (2 mL) followed by slow evaporation of the solvent until a polymer film formed, which was then dried in high vacuum for 2 h. Then ddH2O (1 mL) or buffer (1 mL) was added and the mixture was stirred overnight at room temperature. The cosolvent method consisted of dissolution of copolymer (100 mg) in EtOH (50 μL), which was then added dropwise to 500 μL of ddH2O, followed by dialysis against ddH2O for 24 h with three exchanges of ddH2O. Giant unilamellar vesicles were prepared by the electroformation technique,38 using the Nanion Vesicle Prep Pro setup (Nanion Technologies, Munich, Germany). 50 μL of the polymer solution (4 mg/mL in ethanol) were spread over an ITO-coated glass plate and evaporated in a vacuum chamber for at least 1 h. With an Oring, a chamber was formed around the polymer film, filled with 100 5062
dx.doi.org/10.1021/ma500511r | Macromolecules 2014, 47, 5060−5069
Macromolecules
Article
PMOXA19 was determined as 0.94, thus proving the formation of polymersomes. Further increasing the f PMOXA to 65% induced a change in the architecture of the supramolecular assemblies, and a mixture of small micelles and larger spherical particles with diameter around 70 nm was observed (Supporting Information, Figure S4D−S4F). The aspect of cryo-TEM micrograph (Supporting Information, Figure S4E and S4F) is similar to the very recently reported multicompartment micelles.42,43 However, the resolution of cryo-TEM prevented a more detailed structural analysis. SAXS was also applied to further characterize the spherical particles formed by PDMS65-b-PMOXA32 (Supporting Information, Figure S15). SAXS data of the weakly scattering polymer solution were fitted by a form factor of spherical particles with almost homogeneous density. Owing to the weak scattering a more detailed analysis of the inner structure of the spherical object is not undoubtedly possible. Therefore, the combined cryo-TEM and SAXS did not provide more insight into the internal structure of spherical particles (o.d. ∼ 70 nm). In addition, a smaller ρ value for PDMS65-b-PMOXA32 of 0.82 supports the presence of soft spheres, but without information on an internal structure.44 The transition from polymersomes to micelles was observed for copolymers of poly(ethylene oxide)-block-poly(N,N-diethylaminoethyl methacrylate) (PEO-b-PDEAMA), polystyrene-bpoly(4-vinylpyridine) (PS-b-P4VP), and polystyrene-dendrimer amphiphilic block copolymers with increasing length of hydrophilic blocks.7,25,45 In the PS-b-P4VP system, the polymersomes formed in the range f P4VP 13−24%, which is similar to the range of f PMOXA for our PDMS-b-PMOXA polymer system for polymersome formation. The f PMOXA value mainly determines the shape of the polymeric chains, and therefore influences the size, and morphologies of the selfassembled structures.46,47 The formation of larger spherical particles is presumably caused by increasing the length of the hydrophilic PMOXA, which combines spherical micelles into larger spherical particles. The phase diagram of PDMS-b-PMOXA with a 3.1 kDa PDMS (PDMS39) shows a self-assembly behavior similar to that of the higher molecular mass PDMS (for f PMOXA values >25%) In contrast, this series of copolymers did not generate polymersomes with diameter >100 nm for the whole domain of fPMOXA values (Figure 1B). Polymer aggregates were present during the self-assembly of block copolymers with f PMOXA < 19%, whereas increasing the f PMOXA value to 27% induced the formation of mixtures of polymersomes with diameter
