Structural and Rheological Properties of Temperature-Responsive

Apr 21, 2017 - The turbidity was measured with the aid of a NK60-CPA cloud point ...... Kirschner , C. M.; Anseth , K. S. Hydrogels in Healthcare: Fro...
0 downloads 0 Views 5MB Size
Subscriber access provided by University of Colorado Boulder

Article

Structural and Rheological Properties of TemperatureResponsive Amphiphilic Triblock Copolymers in Aqueous Media Josefine Eilsø Nielsen, Kaizheng Zhu, Sverre Arne Sande, Lubomír Ková#ik, Dušan Cmarko, Kenneth Dahl Knudsen, and Bo Nyström J. Phys. Chem. B, Just Accepted Manuscript • Publication Date (Web): 21 Apr 2017 Downloaded from http://pubs.acs.org on April 23, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Structural and Rheological Properties of TemperatureResponsive Amphiphilic Triblock Copolymers in Aqueous Media

Josefine Eilsø Nielsen,a,b Kaizheng Zhu,b Sverre Arne Sande,a Lubomír Kováčik,c § Dušan Cmarko,c Kenneth D. Knudsen,d Bo Nyström*b

a

School of Pharmacy, Department of Pharmaceutics, University of Oslo, P.O. Box 1068, Blindern, N-0316 Oslo, Norway

b

Department of Chemistry, University of Oslo, P.O. Box 1033, Blindern, N-0315 Oslo, Norway

c

Institute of Biology and Medical Genetics, First Faculty of Medicine, Charles University and General University Hospital in Prague, Albertov 4, Prague, 128 01, Czech Republic

d

Department of Physics, Institute for Energy Technology, P. O. Box 40, N-2027 Kjeller, Norway

§

Current address: Center for Cellular Imaging and NanoAnalytics (C-CINA), Biozentrum, University of Basel, Mattenstrasse 26, 4058 Basel, Switzerland

* Corresponding authors: e-mail [email protected], Phone: +47-22855522.

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT: Thermoresponsive amphiphilic biodegradable block copolymers of the type poly(ε-caprolactone-co-lactide)-poly(ethylene

glycol)-poly(ε-caprolactone-co-lactide)

(PCLA-PEGm-PCLA) have great potential for various biomedical applications. In the present study, we have surveyed the effects of PEG-spacer length (m=1000 and 1500), temperature, and polymer concentration on the self-assembling process to form supramolecular structures in aqueous solutions of the PCLA-PEGm-PCLA copolymer. This copolymer has a lower critical solution temperature, and the cloud point depends on both concentration and PEGlength. Thermoreversible hydrogels are formed in the semidilute regime; the gel windows in the phase diagrams can be tuned by concentration and length of the PEG-spacer. The rheological properties of both dilute and semidilute samples were characterized; especially the sol-to-gel transition was examined. Small-angle neutron scattering (SANS) experiments reveal fundamental structural differences between the two copolymers for both dilute and semidilute samples. The intensity profiles for the copolymer with the long PEG-spacer could be described by a spherical core-shell model over a broad temperature domain, whereas the copolymer with the short hydrophilic spacer forms rod-like species over an extended temperature range. This finding is supported by cryo-TEM images. At temperatures approaching macroscopic phase separation, both copolymers seem to assume extended rodlike structures.

2 ACS Paragon Plus Environment

Page 2 of 60

Page 3 of 60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

1.

INTRODUCTION

Biodegradable and thermogelling polymer systems that can undergo sol-to-gel transition upon a temperature raise have attracted a great deal of attention in recent years for their potential in biomedical applications, including drug delivery, cell therapy, and tissue engineering.1-6 Amphiphilic copolymers, consisting of hydrophilic and hydrophobic groups, have usually the ability to self-assemble to form micelles and intermicellar structures; this may lead to the formation of hydrogels in the semidilute concentration regime. An example of this is aqueous solutions of copolymers composed of poly(lactide-co-glycolide) (PLGA) and poly(ethylene glycol) (PEG) that can form temperature-responsive gelling systems7-9, which have been evaluated in clinical trials for applications in oncology, particularly as matrices for sustained release of paclitaxel.10-13 There are other polymers, composed of ABA triblock copolymers with PEG as the B block and poly(ε-caprolactone-co-lactide) (PCLA) as the A blocks, with similar properties as the PLGA analogs, but in general they have longer degradation times than PLGA-PEG-PLGA-based systems. This may be a great advantage in drug release applications.14,15 Furthermore, the crystallinity of the A blocks is expected to affect the mesoscopic structure and the rheological properties of this copolymer. To prepare and design polymeric nanoparticles for drug delivery and other medical applications, it is crucial to understand the structure of the self-assembled carriers, and how it varies with the copolymer composition, polymer concentration, and external parameters such as temperature or pH. It is necessary to gain this structural insight of the nanostructures to be able to tailor-make carriers for specific applications.16,17 An intricate self-assembling behavior has recently been reported18 for triblock copolymers of the type poly(N-isopropylacrylamide) (PNIPAAM)-b-poly(2-ethyl-2-oxazoline) (PEtOx)-b-PNIPAAM, where both PNIPAAM and PEtOx exhibit lower critical solution temperatures.

3 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

In a recent work19 on thermoresponsive PLGA-PEGm-PLGA block copolymers, which exhibit a reversible temperature-induced sol-to-gel transition at higher polymer concentrations in aqueous solution, we studied the rheological behavior, as well as mesoscopic structural features of triblock copolymers with a changed length of the PEG-spacer at different polymer concentrations and temperatures. The length of the hydrophobic PLGA blocks was kept constant, whereas the molecular weight of the hydrophilic PEG block was altered (m = 1000 and m = 1500 stand for the number average molecular weight) and this change was found to have a strong impact on both the phase behavior of the system and the local structure of the copolymer probed by small-angle neutron scattering (SANS). It was observed in semidilute solutions that a short PEG-block favors gelation at a low temperature, whereas the longer PEG-spacer shifts the gelation point to higher temperature. Dramatic structural changes were observed. In dilute solution, the SANS profiles on the copolymer with short PEG-spacer revealed asymmetric ellipsoid structures, whereas the scattered intensity profile for the copolymer with m=1500 could be modelled by a spherical core-shell model. In the semidilute regime, a less organized structure is formed at the gel point, but an increased temperature led first to appearance of correlation peaks in the SANS profile and to a cylindrical local structure, followed by packing of cylinders for both copolymers at even higher temperatures. The main characteristics are the same for both copolymers, the difference being that the detected features are shifted to higher temperatures for the copolymer with the longer PEGspacer. These findings suggest that intricate structural transformations take place, and they are governed by temperature, polymer concentration, and length of the PEG-spacer. The importance of the length of the PEG-spacer for self-assembling and temperature-induced gelation in PLGA-PEG-PLGA copolymer systems have recently been established.20 Systems based on PLGA-PEG-PLGA copolymers have attracted a great deal of interest for biomedical applications,21 but in some applications longer release times are

4 ACS Paragon Plus Environment

Page 4 of 60

Page 5 of 60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

required and in these cases the PLGA blocks have been replaced by PCLA blocks,22,23 thereby increasing the degradation time and the stability of the gel. In addition, the PCLA component may induce crystalline domains24 that affect the degradation kinetics of the systems in water. It is known25 that crystalline domains are degraded slower than amorphous areas because of limited hydration. The crystalline structure of PCLA is shown to depend not only on the monomer ratio25 but also on the monomer sequence, i.e., its microstructure.26 In the light of these differences between triblock copolymers containing PLGA or PCLA blocks, we became interested in how the rheological and mesoscopic structural features would be affected by replacing the PLGA blocks with PCLA and keeping the PEG-spacer lengths the same. In this work, triblock copolymers of the type PCLA-PEGm-PCLA (m= 1000 and m = 1500 are number average molecular weights) were synthesized by ring opening polymerization. Aqueous solutions of these copolymers were prepared, and the polymers were characterized at different temperatures in both dilute and semidilute aqueous solutions with the aid of various experimental techniques, such as turbidimetry, oscillatory shear, SANS, and cryo-TEM. Since both types of triblock copolymers have great potential for biomedical applications, it is important to survey differences in rheological and structural properties of aqueous systems of these polymers when the PLGA blocks are replaced by PCLA blocks. The aim of this study is to acquire knowledge about how these features are influenced by type of hydrophobic block, temperature, and polymer concentration. This understanding is necessary to be able to tailor-make systems for specific biomedical applications. Although the PCLAPEGm-PCLA and PLGA-PEGm-PLGA systems have several features in common, some important differences are observed, especially when investigating local scale behavior via neutron methods (SANS). For instance, we show in this work that with the same length of the PEG-spacer, the micellar ordering is maintained up to a higher temperature for PCLA-PEGmPCLA than for PLGA-PEGm-PLGA in the semidilute regime.

5 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2.

EXPERIMENTAL METHODS 2.1 Materials. Ɛ-Caprolactone (CL) from Sigma-Aldrich was dried and distilled over

CaH2 under reduced pressure and stored in a refrigerator. D,L-lactide (LA) from SigmaAldrich were recrystallized from ethyl acetate and dried under vacuum and stored at -18 °C before use. Poly(ethylene glycol) (PEG1000 and PEG1500) and stannous 2-ethylhexanoate (stannous octoate, Sn(Oct)2) were purchased from Aldrich-Sigma and used as received without further purification. All other chemicals were reagent-grade and used as obtained. 2.2. Polymer Synthesis. The ABA-type triblock copolymer, poly(ε-caprolactone-coD,L-lactide)-b-Poly(ethylene glycol)-b-poly(ε-caprolactone-co-D,L-lactide), abbreviated as PCLA-PEG-PCLA, was prepared by ring-opening polymerization (ROP) of ε-caprolactone and D,L-lactide with PEG as the initiator and stannous octoate as the catalyst.27-29 Feed ratios of the PEG/CL/LA were used to adjust the composition and molecular weight. The synthetic procedure is outlined below (Figure 1a).

6 ACS Paragon Plus Environment

Page 6 of 60

Page 7 of 60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 1. a) Synthetic route for the preparation of the triblock copolymer PCLA-PEG-PCLA via a ring opening polymerization procedure. b) 1H NMR spectra of PCLA-b-PEG-b-PCLA 7 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

triblock copolymers (CDCl3-d as solvent, 400 MHz). Proton NMR spectra of the polymers with chemical shift and integrals of the peaks are given as supporting information (Figure S1). The detailed synthesis is conducted in the following way: polyethylene glycol (PEG1000) (10 g, 10 mmol) was dissolved in anhydrous toluene (100 mL), and the solvent was azeotropically distilled off to a final volume of 20 mL in the flask to remove the residual water adsorbed by the polymer. When the temperature of the flask had decreased to room temperature, D,L-lactide (12.97 g, 90 mmol) and ε-caprolactone (10.3g, 90 mmol) were added in a mole ratio of 1:1, and the reaction mixture was stirred for another 30 min. Stannous octoate (0.1 mmol, 50 mg in dried toluene (0.5 mL)) was added to the reaction mixtures and stirred at 130 °C for 24 hours in an atmosphere of argon for protection. The flask was then cooled to room temperature; the mixture was dissolved in dichloromethane (50 mL) and it precipitated when it was slowly added into cold diethyl ether. This process of dissolution followed by precipitation was repeated three times. The polymer was further purified by dissolving it in cold water and dialyzing against distilled water at 4 °C for 3 days using a dialysis membrane of regenerated cellulose with a molecular weight cutoff of 1000. The white solid product was finally isolated by lyophilization and kept at -18 °C. The chemical structure and composition of the ABA triblock copolymers were determined by their 1H NMR spectra in CDCl3 solutions containing tetramethylsilane (TMS) as reference at 25 °C (Bruker AVANCE II 400 MHz spectrometer) (Figure 1b). The molar composition of each sample was calculated by comparing the integral area of the PEG methylene signal (1) (δ = 3.65 ppm), the LA single proton (2) (δ = 5.15 ppm), and the CL methylene group (4) (δ = 2.30 ppm). The entire repeating units of CL/LA/EG (2y/4x/n) were estimated to be 9.75/16/22 and 9.95/15.6/34 for PEG1000 and PEG1500 derivative polymers, respectively, based on the calculation that the numbers of repeating units of the ethylene glycol of PEG1000 and PEG1500 are 22 and 34, respectively. Proton NMR spectra of the 8 ACS Paragon Plus Environment

Page 8 of 60

Page 9 of 60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

polymers with chemical shift and integrals of the peaks are given as supporting information (Figure S1). GPC was performed on a Tosoh Eco-SEC dual detection (RI and UV) GPC system coupled to an external Wyatt Technologies miniDAWN Treos multi angle light scatting (MALS) detector. Samples were run in THF at a flow rate of 0.5 mL/min at 35 °C. The column set was one MZ-Gel SDplus linear column (5 µm, 4.6*300 mm). The refractive index increment value (dn/dc =0.063 mL/g) was calculated based on the reported dn/dc values of PCL, PLA and PEG in THF of 0.079, 0.054, and 0.068 mg/mL, respectively.

30,31

Absolute

molecular weights and molecular weight distributions were calculated using the Astra software package. The concentration of the polymer samples is 10.0 mg/mL. Figure 2 shows GPC traces of the copolymers. The elution peaks are symmetric and exhibit no tails at the lower molecular weight side. The characteristic data from the synthesis are given in Table 1.

Figure 2. GPC chromatograms for the two synthesized PCLA-PEG-PCLA triblock copolymers (35 °C, 10 mg/ml with eluent THF and flow rate of 0.5 ml/min).

9 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table 1. Characteristic Parameters from the Synthesis of PCLA-PEG-PCLA Triblock Copolymers Polymers PEG 2CL/2LA/EG Mn (NMR) Mw/Mn (Mn, g/mol) (2y/4x/n) (GPC) ABA1 1000 9.75/16/22 1140-1000-1140 4210/3300 = 1.28 (3280) ABA2 1500 9.95/15.6/34 1180-1500-1180 5430/4090 = 1.32 (3860)

2.3.Turbidimetry. The turbidity was measured with the aid of a NK60-CPA cloud point analyzer made by Phase Technology, Richmond, BC, Canada. The light beam from a light source (AIGaAs, 654 nm) is focused on the sample by means of a lens. 0.15 mL sample is injected by a micropipette on the specific glass plate, which is coated with a metallic layer with a very high reflective index. In fact the plate works as a mirror. The temperature of the sample cell is controlled by a platinum resistance thermometer probe (Peltier elements). To avoid solvent evaporation during a temperature scan, each sample has been covered by the

10 ACS Paragon Plus Environment

Page 10 of 60

Page 11 of 60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

same amount of silicon oil. The scattered intensity signal (S) is detected by the light scattering detector, which is located directly above the investigated sample. The turbidity (τ) and the scattered intensity signal is related to each other by the empirical equation32: τ = 9.0 x 10-9 S3.751. The heating rate in each measurement is 0.2 °C/min over the temperature range of 2560 °C. The temperature at which the first deviation of the scattered intensity from the baseline occurred was taken as the cloud point (CP) of the considered sample. 2.4. Phase Diagram Determination. The phase diagrams for aqueous solutions of the PCLA-PEG-PCLA block copolymers were determined by the tube inverting method.33 Aqueous solutions of the copolymers were prepared at concentrations of 10, 15, 20 and 30 wt% for both copolymers. 1 ml of each solution was transferred into glass tubes. The tubes were sealed and kept in the water bath and heated up slowly from 5 °C to 45 °C. The sample was kept at the measuring temperature for 10 minutes, prior to inspection to ensure equilibrium of the system. The sol-to-gel transition temperature was determined by a flow or no-flow criterion over 60 s. The transition temperature was monitored at an accuracy of better that ±1 °C. 2.5. Rheological Measurements. Oscillatory shear experiments were performed in an Anton Paar-Physica MCR 301 rheometer using a cone-and-plate geometry, with a diameter of 75 mm and a cone angle of 1°, or a cone with 25 mm diameter and a cone angle of 4o. The latter one was used for samples of high concentrations to reduce the amount of consumed copolymer. It was always checked that the results were not affected by the type of cone. The measuring apparatus is equipped with a temperature unit (Peltier plate), which provides an effective temperature control (± 0.05 °C) for an extended time over the studied temperature range. The free surface of solutions was covered with a thin layer of low-viscosity silicone oil to prevent the dehydration of the samples at elevated temperatures. The observed viscosity value is practically unaffected by the oil layer. The rheometer has been calibrated with water

11 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 60

and standard high viscous oil before performing any experiments. The appropriate amplitude sweep for the measurements was chosen to ensure that measurements were conducted in the linear viscoelastic region, where the dynamic storage modulus (G‫ )׳‬and loss modulus (G‫ )׳׳‬are independent of the strain amplitude. The amplitude was held constant at 1%, and the frequency was varied between 0.01-100 Hz. The measurements were performed at every degree from 5 to 45 °C. The sample was held at each temperature for 20 min before performing the measurements to ensure equilibrium in the sample. 2.6. Analysis of Rheological Data. The rheological behavior of an incipient gel can be described by a simple power law, where the dynamic moduli are related as34 G’ = G’’/tan δ = SωnΓ(1 - n) cos δ

(1)

where Γ(1-n) is the gamma function, n is the relaxation exponent, and S is the gel strength parameter, which depends on the cross-linking density and the molecular chain flexibility.34 The phase angle (δ) between stress and strain is independent of frequency (ω) but proportional to the relaxation exponent:

tan δ = G’’/G’ = tan(nπ/2)

(2)

These results suggest that the following scaling relation can describe the incipient gel:

G′(ω) ∝ G′′(ω) ∝ ω n

(3)

Muthukumar elaborated a theoretical model,35 based on the hypothesis that strand length variations between crosslinks in the incipient gel network give rise to alterations of the excluded volume interactions, to rationalize values of the relaxation exponent n in the whole accessible region (019 wt%; PCLA-PEG(1500)-PCLA), the solutions undergo a reversible sol-to-gel transition as the temperature is increased. It should be noted that the gel region extends to higher values for the copolymer with long PEG-spacer (e.g., at a concentration of 22 wt% it extends to 35 °C and 28 °C for the copolymer with long and short PEG-spacer, respectively); above these temperatures the gel becomes more turbid and gradually macroscopic phase separation takes place. The fact that the gel-concentration is lower and the gel area is narrower for the PCLA-PEG(1000)-PCLA copolymer than for the one with long PEG-spacer can be rationalized in the following way. The gel network is formed in the semidilute concentration regime, and it is controlled by a delicate balance between swelling and network connectivity. If swelling is the predominant feature and the physical crosslinking zones are in deficiency, a transient network is formed but not a gel network. On the other hand, if the connectivity through large hydrophobic patches is the prevailing feature, phase separation may occur and no gel is formed. For the present polymer systems, it is likely that the hydrophobic microdomains in the network to a large extent provide the connectivity. In the light of this, it is reasonable to expect that the PCLAPEG(1000)-PCLA system should have a lower gelation-concentration, incipient phase separation should occur at a lower temperature, and narrower gel-phase region than for the

18 ACS Paragon Plus Environment

Page 18 of 60

Page 19 of 60

PCLA-PEG(1500)-PCLA system (see Figure 4). Since the former copolymer is more hydrophobic it should be more inclined to form hydrophobic microdomains and thereby generate the necessary network connectivity for the gel earlier than for the less hydrophobic copolymer.19 In addition, upon an increase in temperature it is expected that the more hydrophobic polymer should develop an excess of cross-liking zones and an abnormal high connectivity at a lower temperature than for the copolymer with a long PEG-spacer. The high connectivity will eventually lead to macroscopic phase separation. In a previous study44 on PCLA-PEG-PCLA samples, it was argued that gelation is due to strong hydrophobic interactions at high temperature, and it was found that the temperature domain at which gels formed was affected mainly by the ratio of hydrophobic to hydrophilic blocks. 45 PCLA-PEG(1500)-PCLA PCLA-PEG(1000)-PCLA

40 35

Temperature (oC)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

30

SOL (Phase separation)

25 20

GEL

15 10 SOL

5 0 10

12

14

16

18

20

22

24

26

28

30

32

Weight percentage

Figure 4. Phase-diagrams for the two triblock copolymers: PCLA-PEG(1000)-PCLA (black color) and PCLA-PEG(1500)-PCLA (red color) in aqueous media. The lines are added as guide for the eye. The inset images illustrate sol-phase, gel-phase, and phase separation (high temperatures) for a 20 wt% sample of the PCLA-PEG(1000)-PCLA copolymer.

19 ACS Paragon Plus Environment

The Journal of Physical Chemistry

In Figure 5, the dynamic storage (G’) and loss (G’’) moduli are plotted against temperature for the two triblock copolymers at a polymer concentration of 20 wt% and at a low constant frequency of ω = 0.6 s-1. In the approximate temperature interval 11-19 °C, G’>G’’ for the PCLA-PEG(1000)-PCLA system and the elastic response dominates,

which

indicates that the gel is fully developed in this temperature range. At higher temperatures G’’>G’ and the viscous response prevails and the gel is gradually transformed to a sol. For the PCLA-PEG(1500)-PCLA system, G’>G’’ in the temperature domain from 7 °C to 22 °C and this wider region is anticipated because for this copolymer the gel region is broader.

104 20 wt% PCLA-PEG(1000)-PCLA:

103

G', G'' (Pa)

102 101 100 10-1 10-2

G' G''

-3

10

4

6

8

10

12

14

16

18

20

22

24

Temperature (°C)

104 20 wt% PCLA-PEG(1500)-PCLA:

103 102

G', G'' (Pa)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 60

101 100 10-1 10-2

G' G''

10-3 4

6

8

10

12

14

16

18

20

22

24

Temperature (°C)

Figure 5. Temperature dependences of the storage modulus (G’) and the loss modulus (G’’) for 20 wt% aqueous samples of PCLA-PEG(1000)-PCLA and PCLA-PEG(1500)-PCLA at a low frequency of 0.6 s-1. 20 ACS Paragon Plus Environment

Page 21 of 60

It is interesting to note that at higher temperatures the values of G’ for the copolymer with the longer PEG-spacer are almost two order of magnitude larger (1000 Pa) than the corresponding values for the other copolymer. In addition to dehydration of PEG and enhanced hydrophobic interactions at elevated temperatures, the reason for this momentous difference between the two copolymers can probably be attributed to the formation of intermicellar bridges22,23,27,45 for the copolymer with the long PEG-spacer. It is obvious that this change in the length of the PEG-spacer leads to a substantial mechanical strengthening of the network. To improve the strength of the gel, PCLA-PEG-PCLA has been end-capped with peptide45 and hydroxyl hexanoyl end groups22, as well as acetyl-capped.23 These modifications led to stronger gels, but only maximum values of approximately 100 Pa of G’ have been reported22,23,45 for the PCLA-PEG-PCLA copolymer at a polymer concentration of 25 wt% in aqueous media. a)

103 Dynamic viscosity (Pas)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

b)

102 101 100 10-1 10-2 10wt% 20wt% 30wt%

10-3 10-4

0

10

20 30 Temperature (°C)

40

10wt% 20wt% 30wt%

50 0

10

20 30 Temperature (°C)

40

50

Figure 6. Temperature dependences of the dynamic viscosity for a) PCLA-PEG(1000)-PCLA and b) PCLA-PEG(1500)-PCLA at the concentrations indicated and at a constant low frequency of 0.1 Hz.

21 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 60

Temperature dependences of the dynamic viscosity for the two copolymers at different concentrations are depicted in Figure 6. The general trend of the dynamic viscosity (η’=G’’/ω) for the two copolymers is that it rises with increasing polymer concentration; the impact of concentration on the dynamic viscosity (η´) is more pronounced for the copolymers with the longer PEG-spacer. As mentioned above, this is ascribed to the presence of intermicellar bridges for the copolymer with the longer PEG-spacer and this effect is augmented as the concentration increases. The most conspicuous difference in dynamic viscosity behavior between the two copolymers is at a concentration of 10 wt%, where η´ for the copolymer with the short PEG-spacer passes through a pronounced maximum, whereas the dynamic viscosity for the other copolymer only exhibits a modest decrease with increasing temperature; at the highest temperatures a slight upturn in the dynamic viscosity is detected (Figure 6). This concentration is located in the dilute regime, where we have individual micelles. In the case of the more hydrophobic polymer (short PEG block), a temperature rise induces development of intermicellar complexes, which result in a significant increase in the dynamic viscosity. This feature is ascribed to enhanced sticking probability of the species at elevated temperatures and the formation of loose intermicellar structures as a result of hydrophobic interactions and dehydration as the temperature rises. The formed agglomerates consist of a number of interconnected micelles and as the temperature rises the many hydrophobic segments inside the clusters will be close-packed to avoid water exposure and this process may progressively lead to break-up of the clusters because of contraction of the subunits. At 10 wt% solution of the PCLA-PEG(1500)-PCLA copolymer,

η´ decreases

moderately up to about 40 °C, and thereafter an upturn in η´ is visible. This finding suggests that over a fairly wide temperature range we have contraction of the micelles and at higher temperatures intermicellar structures are formed due to enhanced sticking probability. This is

22 ACS Paragon Plus Environment

Page 23 of 60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

consistent with the turbidity increase observed at high temperatures for the PCLAPEG(1500)-PCLA copolymer (see Figure 3). For the PCLA-PEG(1000)-PCLA copolymer at 20 wt% (semidilute concentration regime), the viscosity maximum is less pronounced and the drop of the viscosity at elevated temperatures can probably be ascribed to dehydration and disruption of the gel-network as phase separation is approached at higher temperatures (see Figures 3 and 4). At the same conditions for the PCLA-PEG(1500)-PCLA sample, η´ rises as the gel is evolved up to ca. 15 °C; thereafter η´ is virtually constant up to about 40 °C and η´ falls off as the temperature comes close to the phase separation temperature. At the highest concentration (30 wt%), the dynamic viscosity for the copolymer with the short PEG-spacer falls off already at 15 °C as the turbidity increases strongly and the network is fragmented as phase separation is advanced and η´ falls off considerably. For the more hydrophilic copolymer, the drop of η´ is less pronounced and it is shifted to a higher temperature (≈30 oC). We note the strong decrease of η´ for the PCLA-PEG(1000)-PCLA system at high temperatures. This is probably caused by macroscopic phase separation, which may lead to an artefact in the rheology measurements due to loss of contact between the cone and the plate of the rheometer and the gel phase. Above some basic rheological properties of the two copolymers have been characterized, and the approximate gelation window for polymer samples of various concentrations has been established by using the tube inverting method. In this context it is interesting to investigate whether it is possible to utilize the classical approach by Winter and Chambon46 to determine the incipient gelation temperature. In this method, the gel point can be determined by observation of a frequency-independent value of tan δ (=G´´/G´) obtained from a multi-frequency plot of tan δ versus temperature (see Figure 7a). An alternative method (see the insets) to determine the gelation temperature is by plotting the “apparent”

23 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

viscoelastic exponents n' and n'' (G'~ ωn', G''~ ωn") calculated from the frequency dependence of G' and G'' at different temperatures and observing a crossover where n'=n''=n.47 It is evident from Figure 7a (30 wt% of PCLA-PEG(1000)-PCLA) that these methods yield the same gelation temperature of 7 °C. The general feature is that the loss tangent (tan δ) is frequency dependent and decreases during the gel formation, indicating that the solutions become more and more elastic. At the gel point, the G’ and G’’ curves in a log-log plot become parallel and power laws in frequency (see insets) are found as predicted by the Winter and Chambon model.

24 ACS Paragon Plus Environment

Page 24 of 60

1 0.9 a)0.24 0.8 tan δ

0.20

0.6

B)

7 °C:

Storage Modulus Loss Modulus

n', n''

0.22

0.7

1000

A)

Critical gel point T = 7°C

n' n''

0.18

100

0.5

6

8

10

0.1

1

Critical gel point T = 7°C

0.4 0,628 rad/s 0,996 rad/s 1,58 rad/s 2,5 rad/s

0.3 4

6

8 Temperature (°C)

10

12

0.6

b)

2

10 0.5

14°C:

Storage Modulus Loss Modulus

n' n'' G', G''(Pa)

n'. n''

1.5

10

Angular Frequency (rad/s)

Temperature (°C)

0.4

tan δ

0.3

12

13

14

15

1

10

1

1.0

10

Angular Frequency (rad/s)

0,628 0,996 1,58 2,5 3,96

0.5 12

13

14 15 16 Temperature (°C)

c)

17

18

0,628 rad/s 0,996 rad/s 1,58 rad/s 2,5 rad/s

0

10

tan δ

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

G', G'' (Pa)

Page 25 of 60

10-1 5

10

15

20

Temperature (°C)

Figure 7. Viscoelastic loss tangent (tan δ) as a function of temperature for a) 30 wt% of PCLA-PEG(1000)-PCLA and b) 20 wt% of PCLA-PEG(1000)-PCLA at the frequencies in Hz indicated. The inset plots to the left depict changes of the apparent exponents n’ for the storage and n’’ for the loss moduli during the course of gelation. The inset plots to the right

25 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

show the power law behavior of the dynamic moduli at the gel point temperatures. c) Tan δ versus temperature for 20 wt% of PCLA-PEG(1500)-PCLA. For the 30 wt% PCLA-PEG(1000)-PCLA sample, the sol-to-gel transition is distinct and easily identified from the plots in Figure 7a. However, in the case of the 20 wt% PCLAPEG(1000)-PCLA system (Figure 7b) it is more intricate to establish the gel point from the two methods; it seems from the tan δ plot that 14 °C is the gel point, whereas from the n’ and n’’ plot it is difficult to conclude. This gelation temperature is much higher than that (7 °C) observed for the high polymer concentration. A higher value of the gel point is expected for a lower polymer concentration, because in this case a higher temperature is required to establish the hydrophobic microdomains that are necessary for the connectivity of the gel-network.19 The value of the exponent n is close to 0.3 at 20 wt% and n≈ 0.2 for 30 wt%. In terms of fractal dimension (see eq 4), this suggests df=2.2 and df=2.3, respectively. This indicates a more compact network at the highest polymer concentration, as expected since more space in the network should be occupied. Figure 7c shows a plot of tan δ versus temperature at various frequencies for 20 wt% of PCLA-PEG(1500)-PCLA and it is not possible to accurately determine the sol-to-gel transition. By using the tube inverting method, the incipient gel temperature is determined to be ca. 12 °C. This temperature is lower than the corresponding one for the PCLA-PEG(1000)PCLA system, in spite of that the former copolymer is less hydrophobic. The reason for this anomaly may be that the long PEG spacer facilitates intermicellar bridging and this contributes to establish the network connectivity. The problem to determine the gel point from the Winter-Chambon method has also been noticed for temperature-induced Pluronic (PEO-PPO-PEO) gels.48,49 This problem is usually not observed for chemically cross-linked incipient gels,50,51 where the formation of the incipient gel is rather abrupt. The conjecture is that when the transition from sol-to-gel is a 26 ACS Paragon Plus Environment

Page 26 of 60

Page 27 of 60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

smooth and progressive process, the rheological parameters are changed gradually and there is not an abrupt gel point. Gels formed in this way are expected to be rather fragile. 3.3. Small Angle Neutron Scattering Experiments. To survey temperature-induced structural changes on a mesoscopic length scale, SANS experiments were conducted on dilute and semidilute solutions of the two triblock copolymers. To be able to follow the discussion of the results, a brief summary of the employed models will be presented. 3.3.1. Modelling of the SANS Data. The SANS data were fitted to different models as described in the text below. The core-shell spherical model has a form factor P(q) given by

 =

  ρ ρ  







+

 ρ ρ   !

(5)



where Vs is the particle volume including the outer shell, Vc is the volume of the core, Rs is the radius of the shell, and Rc is the radius of the core, thus Rs = Rc + d, where d is the thickness of the shell. The parameter ρc is the scattering length density of the core, ρs is the scattering length density of the shell, and ρsolv is the scattering length density of the solvent. j1(x) = (sin x - x cos x)/ x2, and scale notation is a scale factor proportional to the sample concentration. The core-shell cylinder model that was also employed in this study, has a form factor P(q) described by

 =

 

$

"&% f ! q, α sinα dα

(6)

where it has been assumed an overall random orientation of cylinders, and where

- , α = 2/0 − /2 30 4&  56789

:  ;