Selective Tuning of the Self-Assembly and ... - ACS Publications

Article pubs.acs.org/Langmuir

Selective Tuning of the Self-Assembly and Gelation of a Hydrophilic Poloxamine by Cyclodextrins Gustavo González-Gaitano,*,† Marcelo A. da Silva,‡ Aurel Radulescu,§ and Cécile A. Dreiss*,‡ †

Departamento de Química y Edafología, Universidad de Navarra, 31080 Pamplona, Spain Institute of Pharmaceutical Science, King’s College London, Franklin-Wilkins Building, 150 Stamford Street, London SE1 9NH, U.K. § Jülich Center for Neutron Science, JCNS Outstation at MLZ, Forschungszentrum Jülich GmbH, Lichtenbergstraße 1, 85747 Garching, Germany ‡

S Supporting Information *

ABSTRACT: Complexes formed between cyclodextrins (CDs) and polymers pseudopolyrotaxanes (PPRs) - are the starting point of a multitude of supramolecular structures, which are proposed for a wide range of biomedical and technological applications. In this work, we investigate the complexation of a range of cyclodextrins with Tetronic T1307, a four-arm block copolymer of poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) with a pHresponsive central ethylene diamine spacer, and its impact on micellization and the sol−gel transition. At low concentrations, small-angle neutron scattering (SANS) combined with dynamic light scattering (DLS) measurements show the presence of spherical micelles with a highly hydrated shell and a dehydrated core. Increasing the temperature leads to more compact micelles and larger aggregation numbers, whereas acidic conditions induce a shrinking of the micelles, with fewer unimers per micelle and a more hydrated corona. At high concentrations, T1307 undergoes a sol−gel transition, which is suppressed at pH below the pKa,1 (4.6). SANS data analysis reveals that the gels result from a random packing of the micelles, which have an increasing aggregation number and increasingly dehydrated shell and hydrated core with the temperature. Native CDs (α, β, γCD) can complex T1307, resulting in the precipitation of a PPR. Instead, modified CDs compete with micellization to an extent that is critically dependent on the nature of the substitution. 1H and ROESY NMR combined with SANS demonstrate that dimethylated β-CD can thread onto the polymer, preferentially binding to the PO units, thus hindering self-aggregation by solubilizing the hydrophobic block. The various CDs are able to modulate the onset of gelation and the extent of the gel phase, and the effect correlates with the ability of the CDs to disrupt the micelles, with the exception of a sulfated sodium salt of β-CD, which, while not affecting the CMT, is able to fully suppress the gel phase.

■

INTRODUCTION Supramolecular chemistry has been a strong focus of research in soft matter over the past couple of decades. Intricate structures with bespoke functionality and responsive behavior can be crafted by exploiting noncovalent interactions.1 For this purpose, host−guest chemistry, based on the favorable localization of a guest molecule within the cavity of a host, is a particularly versatile tool for creating soft, functional materials. Inclusion complexes between polymers and cyclodextrins (cyclic oligosaccharides of glucopyranose units) based on this interaction, so-called pseudopolyrotaxanes (PPR), are at the origin of dynamic supramolecular structures,2 which are finding increasing uses in biomedical and technological fields,3−5 based on specific features imparted by the threading of the CD onto the polymer chain, such as improved solubility, nanoscale organization, multivalency, and the sol−gel transition.1 In this work, we investigate the impact of both native and modified cyclodextrins (CD) on the self-assembly and gel behavior of poly(ethylene oxide)-poly(propylene oxide) block © 2015 American Chemical Society

copolymers, so-called Tetronics, or poloxamines (Scheme 1, SI), and the formation or otherwise of PPRs and the interactions responsible for the complexes, which is still a subject of debate in the literature.6−8 We focus here on large, hydrophilic T1307 (Mw 18 000 g·mol−1, HLB > 24), which presents a sol−gel transition at high concentrations and temperatures, a feature of interest for the design of injectable thermosensitive gels, which could act as a sustained drug release depot in situ. Tetronics are amphiphilic block copolymers, presenting an original X shape, where each of the four arms is made of a PPO and a PEO block connected by a central ethylene diamine spacer. The HLB of the molecule can be varied by tuning the length of the PEO and PPO blocks on each arm, leading to a rich phase behavior, which, however, has been scarcely studied to date,9,10 despite the proven potential of Tetronics for a Received: March 24, 2015 Revised: May 1, 2015 Published: May 4, 2015 5645

DOI: 10.1021/acs.langmuir.5b01081 Langmuir 2015, 31, 5645−5655

Article

Langmuir variety of applications, including drug and gene delivery11−13 and tissue engineering.14 A clear advantage of Tetronics over their more well-known linear analogues, Pluronics, is derived from the possibility of protonating the central amino groups, which bestows responsiveness to the assembly process, making drug loading and release pH-sensitive and imparting the possibility of site-specific delivery. In addition and similarly to Pluronics, Tetronics have been shown to inhibit the activity of transmembrane proteins belonging to the ATP-binding cassette protein superfamily (ABC), which pumps drugs out of cells and is overexpressed in cancer cells, thus conferring multiple drug resistance (MDR).15 These results highlight the potential of Tetronics both as nanocarriers and ABC inhibitors in MDR tumors and infections that involve the activity of these efflux transporters. Because of their toruslike structure, relative hydrophobic cavity, and capacity to form hydrogen bonds, cyclodextrins are known to thread onto polymer chains, with α-CD binding preferentially to PEO chains and β-CD binding to PPO blocks.16,17 Native CDs induce the formation of insoluble PPRs or supramolecular gels mainly by reducing the hydrogen bonds between the polymer and water, which have been proposed for the controlled release of drugs18 and proteins.19,20 However, the insolubility of PPRs is a limitation for biomedical applications. Soluble PPRs may be obtained instead from modified CDs (where one or all of the hydroxyl groups around the CD rim are substituted by other chemical groups), but their capacity to form PPRs is far less understood and studied; the outcome of the interaction, however, appears to be highly dependent on the substitution pattern.21,22 In specific cases, CDs have been shown to induce full micellar rupture,6,23 which may be envisaged as a trigger for the controlled release of a loaded cargo.18−20,24 PPRs based on poloxamines are largely unreported, and studies to date have been mainly limited to the native (nonsubstituted) CDs.18−20,25 In this work, we systematically explore the effect of a range of substituted CDs on the micellization process of Tetronic T1307 using a combination of techniques (SANS, DLS, NMR, and rheology) to study the detailed structure of the micelles and complexes and elucidate the key interactions involved as a function of temperature, pH, and CD substitution pattern. We show in particular that substituted CDs have a strong impact in modulating the extent and the onset of the gel phase of Tetronics. Overall, our results provide new insights into the structure and mechanism of complexation between modified CDs and Tetronics and their impact on gelation; these findings provide a robust rationale for the development of pH- and temperatureresponsive materials for biomedical applications.

■

each CD are depicted in SI Scheme 1. Unless otherwise stated, the concentrations are expressed in weight percentage. Dynamic Light Scattering (DLS). Size distributions were obtained with a photon correlation spectrometer (Malvern Zetasizer Nano) with a laser wavelength of 633 nm. The samples were filtered prior to the measurements with 0.22 μm Millex syringe PVDF filters on semimicro glass cells, and the temperature of the sample was controlled to 0.1 °C accuracy by the built-in Peltier in the cell compartment. The viscosities and refractive indices of the solvent at different temperatures were taken into account to obtain the particle size distribution from the analysis of the autocorrelation function, which was determined with the Zetasizer software in high-resolution mode to distinguish overlapping distributions. Small-Angle Neutron Scattering (SANS). Small-angle neutron scattering (SANS) experiments at natural pH were performed on the KWS-2 instrument at the Jülich Centre for Neutron Science (JCNS), Munich, Germany. An incidental wavelength of 4.97 Å was used with detector distances of 1.7 and 7.6 m to cover the q range from 0.007 to 0.3 Å−1. The temperature was controlled by a Peltier system with an accuracy of 0.1 °C. All samples were measured in quartz cells (Hellma) with a path length of 2 mm using D2O as the solvent (either Aldrich >99.9% in D or Armar chemicals >99.8% in D). The experiments on the poloxamine at acidic pH were carried out on the LOQ instrument at the ISIS pulsed neutron source (ISIS, Rutherford-Appleton Laboratory, STFC, Didcot, Oxford). LOQ uses incident wavelengths from 2.2 to 10.0 Å, sorted by time of flight, with a fixed sample− detector distance of 4.1 m, which provided a range of scattering vectors (q) from 0.009 to 0.29 Å−1. The samples were prepared by dissolving the poloxamine or its mixtures with CDs in D2O. They were placed in clean disc-shaped quartz cells (Hellma) of 2 mm path lengths, and the measurements were carried out at different temperatures from 20 to 65 °C. All scattering data were first normalized for sample transmission and then background-corrected using a quartz cell filled with D2O (this process also removes the inherent instrumental background arising from vacuum, windows, etc.) and finally corrected for the linearity and efficiency of the detector response using an instrument-specific software package. The data were then converted to the differential scattering cross sections (in absolute units of cm−1) using the standard procedures at ISIS.26 Analysis of the SANS Data. SANS curves were fitted using the SasView 3.0.0 software.27 Scattering curves from T1307 in its unimer form were fitted with a four-arm star-shaped polymer model. Micelles were fitted to a core−shell sphere (CSS) model combined with a hardsphere structure factor. At intermediate temperatures, where both unimers and micelles coexist, the star-shaped polymer model was combined with a CSS model, weighted out by their respective volume fractions (which were left as floating parameters). Nagg, the aggregation number of the micelles, was obtained from the value of the scattering length density (sld) of both the core and the shell. At low concentration of the poloxamine, leaving the sld of the core as a free parameter yields values close to that of pure PO (3.44 × 10−7 Å−2), which suggests a very low penetration of solvent; this parameter was therefore fixed. Instead, the shell is extensively hydrated, and the level of hydration can be estimated from the fitted value of ρshell, the sld of the PEO shell. The volume fraction of solvent in the corona, xsolv, is related to the sld of the shell, PEO block, and D2O by

MATERIALS AND METHODS

xsolv =

Materials. Tetronic 1307 (T1307) was a gift from BASF. The reported composition per arm is 72 EO and 23 PO, with an 18 000 g· mol−1 average molecular weight. The following CDs were used: αcyclodextrin (α-CD, Fluka, 98%), β-cyclodextrin (β-CD, Sigma 98%), γ-cyclodextrin (γ-CD, Aldrich, 90%), heptakis(2,6-di-O-methyl)-βcyclodextrin (DIMEB, Sigma, 98%), heptakis(2,3,6-tri-O-methyl)-βcyclodextrin (TRIMEB, Sigma, 90%), (2-hydroxyethyl)-β-cyclodextrin (HEBCD, Aldrich, 0.7 molar substitution), (2-hydroxypropyl)-αcyclodextrin (HPACD, Aldrich, 0.6 molar substitution), a sulfated sodium salt of β-cyclodextrin (SBCD, Aldrich, 12−15 substituents per molecule), and randomly methylated β-cyclodextrin (RAMEB, Aldrich, 1.6−2.0 mols CH3 per glucose unit). The repeating units of

ρshell − ρEO ρD O − ρEO

(1)

2

The number of water molecules in the shell is obtained from

n D2O = xsolv

Vshell vD2O

(2)

where vD2O is the volume of a molecule of solvent. The number of water molecules per EO group, nsolv/EO, can then be obtained from eq 2, and the value of the aggregation number, Nagg, is obtained from the hydration of the shell and the structural parameters of the core−shell 5646

DOI: 10.1021/acs.langmuir.5b01081 Langmuir 2015, 31, 5645−5655

Article

Langmuir model. Provided that the amount of water inside the core is negligible, the volume of the micelle is

Vm = Naggvs + xsolvVshell

monoprotonated ones, respectively. As a result, the pH also impacts the extent and onset of the gel phase: at pH 6, a 20% solution forms a gel over a slightly narrower temperature range (40 to 55 °C) than at natural pH. At pH 3, however, the gel phase appears only at 30% and 80 °C, suggesting that charge repulsion between the monomers hinders micelle formation and their subsequent coalescence into a gel phase. Under dilute conditions (1%), unimers are the only species present up to 25 °C, with a hydrodynamic radius of 5.5 nm, as obtained by DLS (Figure 2A), along with a broad peak at sizes larger than the filter pore cutoff (0.22 μm). This broad peak is a typical feature of poloxamers and poloxamines in water, attributed to large clusters of hydrophobic impurities in the form of mono or diblock copolymers, having a negligible contribution either in mass or number, that dissolve in the micelles when the CMT is reached. Micelles start forming above 25 °C, as detected by a second, broad distribution centered at 15 nm at 30 °C. Both distributions merge at 35 °C, and the size (about 9 nm in hydrodynamic radius) remains constant with temperature, signaling a complete transition from unimers to micelles. One notable feature is the relatively small size of the micelles compared to that of the unimers. In comparison, T904, which has a comparable, slightly shorter, hydrophobic block (17 versus 23) but a much shorter EO block (15 versus 72 per arm), forms micelles of 5.8 nm radius, and its unimers are 2.1 nm (ref 28). The ratio of micelle to unimer hydrodynamic radius (Rmic/Runi) is much lower for T1307, suggesting that micelles contain fewer unimers and may be more loosely aggregated. This behavior has been observed in the related family of Pluronic copolymers, when comparing Pluronics of high HLB, for instance, F108 (56 POs, 295 EOs, aggregation number of 13), with more hydrophobic ones of lower molecular weight, such as P85 (38 POs, 51 EOs per block, aggregation number of 57, ref 29). Further structural details of the micelles can be obtained from small-angle neutron scattering (SANS) measurements, using the DLS results to identify the composition of the systems at each temperature (i.e., micelles vs unimers). Figure 3A shows an increase in intensity of the curves with temperature, reflecting the micellization process. We consider here that a four-arm star-shaped polymer is a realistic model for a poloxamine. The form factor is given by30

(3)

where vs is the volume of a surfactant molecule. Nagg can be extracted by introducing into the equation the volume fraction of solvent in the shell deduced from eq 1. At high concentrations of poloxamine (20%), the core becomes hydrated and Nagg was calculated from a modified version of eq 3, which takes into account the hydration of the core: s c Vm = Naggvs + xsolv Vshell + xsolv Vcore

(4) 1

Nuclear Magnetic Resonance. Monodimensional H NMR spectra were recorded on a Bruker Advance 400 MHz spectrometer. For the 2D-ROESY experiments, a Bruker AVIII 700 was used. Samples were prepared by dissolving the solutes in D2O. The residual HDO signal was used as the reference in the spectra. Rheology. Small-amplitude oscillatory experiments were performed on a dynamic strain-controlled rheometer (ARES, TA Instruments) using parallel-plates geometry (25 mm), with a temperature-controlling Peltier unit and a solvent trap. All solutions were prepared by weighing the necessary amounts of solutes and solvent followed by vigorous stirring combined with cooling cycles (4 °C) until their complete dissolution. The samples were left to rest at least 1 day at room temperature after preparation before conducting the rheological measurements. After loading, a thin layer of lowviscosity paraffin oil was added to the geometric edge in order to prevent evaporation. Samples were allowed to rest for a few minutes before the start of the experiments to ensure dissipation of any preshearing due to manipulation and loading. Temperature sweeps at a frequency of 6.28 rad·s−1 and 0.5% strain amplitude, within the limit of the linear viscoelastic range as measured by strain amplitude experiments, were conducted, at a heating rate of 2 °C/min, to cover the temperature range from 20 to 80 °C.

■

RESULTS AND DISCUSSION Self-Aggregation of T1307 and Phase Behavior. Unimers and Micelles. Solubilization of T1307 in water leads to clear solutions which, depending on the temperature, pH, and concentration, undergo a sol−gel transition (Figure 1). For

I(q) =

⎫ f−1 2 ⎧ ⎨v − 1 + exp( −v) + [1 − exp(−v)]2 ⎬ 2⎩ ⎭ 2 fv (5)

where f is the number of arms and v=

Figure 1. Phase behavior of T1307 in aqueous solutions. ○ solution; □ viscous solution; ● gel.

u 2f and u = R g 2q2 (3f − 2)

From the fits, it can be seen how the radius of gyration of the unimers, Rg, increases with temperature, from 32.0 Å at 20 °C to 42.4 Å at 30 °C (where unimers coexist with micelles). At higher temperatures (40 °C and above) only micelles are present, with overall radii that match those determined by DLS (Figure 2). In addition, SANS data provide indirect information on solvent penetration from the value of the scattering length density (sld) of both the core and the shell (cf. Materials and Methods). Results from the fits (Table 1) show that increasing temperatures dehydrate the shell, as the number of solvent molecules per EO unit diminishes while micellar dimensions as

example, at natural pH, a 20% T1307 sample is a gel from 35 to 55 °C. The gel region expands with concentration: at 30 wt %, the gel phase is present from rt up to 80 °C (maximum temperature measured). The self-aggregation ability of poloxamines is known to depend on pH, as the amino groups of the middle block are susceptible to protonation. The reported pKa values for T1307 are pKa1 = 4.6 and pKa2 = 7.8 (ref 9). The natural pH of a 1% aqueous solution is 8.3, which implies that the dominant species is the neutral poloxamine, with a small contribution from the monoprotonated form, whereas at pH 3 and 6 the dominant forms are the diprotonated and 5647

DOI: 10.1021/acs.langmuir.5b01081 Langmuir 2015, 31, 5645−5655

Article

Langmuir

Figure 2. Intensity size distribution as a function of temperature obtained by DLS of a 1% aqueous solution of T1307 at (A) pH 8.3 and (B) pH 2.8.

reflects the high concentration of particles, and the value of ρcore reveals a certain degree of solvent penetration into the micellar core. However, at 36 and 50 °C (gel phase) both the sld of the core and its radius increase considerably while the thickness of the shell is reduced, thus switching their relative importance. All of these changes imply a compression and dehydration of the PEO chains as well as the penetration of solvent molecules into the micellar core. Results from the fits (Table 2) reflect a transfer of the solvent from the core to the shell with temperature, accompanied by an increase in Nagg (calculated from eq 4, cf. Materials and Methods). To the best of our knowledge, the only published SANS study on Tetronic gels has been performed on a reverse poloxamine32 (RED 9040, with nominal 16 EO and 19 PO per arm and Mw = 7200 g· mol−1). Although RED 9040 does not form a gel in the strict rheological sense but rather very viscous solutions, an intense peak reflecting strong correlations between the particles was detected in the q region around 0.07−0.1 Å−1, and the fit to a core−shell model with a hard-sphere structure factor gave qualitatively similar results and trends, including the micellar size going through a maximum value with temperature. An interesting result is the constant value of the volume fraction returned by the fits at 0.63 (it diminishes slightly at 65 °C, where the gel is on the verge of becoming a solution). The theoretical volume fraction is 0.74 for a compact packing of spheres (either cubic or hexagonal), 0.68 for a bcc, and 0.52 for a simple cubic arrangement. In the case of packed micelles, this value could be higher, as a certain overlap of spheres sharing hydrophilic shells can take place. In recent studies on Pluronic F127 gels, the scattering pattern clearly indicates the presence of a paracrystal21,31,33,34 at 18%, with bcc packing being reported as the most plausible type of packing.35 In our case, the failure to fit the data to any of the typical sphere packings, together with the estimated fraction volume of 0.63, seems to indicate that the arrangement is random close-packed, in which the density limit is 63.4% for monodisperse spherical objects.36

well as Nagg increase. These data taken together suggest that temperature promotes a more compact structure of the micelles. At acidic pH (pH 1.9), the size of the unimers at 20 °C is virtually identical to that at 30 °C, (Rg = 34.6 and 34.1 Å, respectively, Figure 2B) and slightly larger than at natural pH (32.0 Å). The positive charge on the central group, creating repulsions between the hydrophobic residues, may be responsible for this slight expansion of the unimers. At 40 and 50 °C, micelles are considerably smaller than at natural pH, with Nagg dropping to 4 and 8 at 20 and 30 °C, respectively (Table 1), a finding also reflected in the DLS size distributions (Figure 2B). The PEO blocks are also more hydrated than at natural pH, suggesting overall much looser micellar structures. González-López et al.9 obtained an aggregation number of 5 from static light scattering under similar conditions (10 mmol· L−1 of HCl). It is interesting that a change to very acidic pH, envisaged as a trigger for the release of a cargo from the interior of them, does not totally destroy the micelles, which would therefore allow delivery in a stepwise fashion. Gel Phase. Raising the temperature and concentration induces the formation of a clear gel phase. SANS curves from 20% solutions in the viscous (20 °C) to the gel phase (up to 65 °C) show the occurrence of a strong correlation peak around q = 0.05 Å−1 (Figure 3C). At 20 °C, the fit to a core−shell model produces micelles with a 26 Å core radius and shell thickness of 34 Å (Table 2). At 36 °C, the solution turns into a gel, producing a high-intensity peak, as is often observed in physical gels of packed micelles.31 The scattering curves, however, could not be fitted to a simple cubic, bcc, or fcc para-crystal model, even when introducing fluctuations as a Lorentzian term (as described for Pluronic gels21). Very good results were instead obtained using the same core−shell model combined with a hard-sphere structure factor, leaving ρcore as a free parameter to account for the hydration of the core (Table 2). At 20 °C, where 20% solutions are visibly more viscous than water but no gelation has occurred, the high volume fraction 5648

DOI: 10.1021/acs.langmuir.5b01081 Langmuir 2015, 31, 5645−5655

Article

Langmuir

Figure 3. SANS curves from 1% T1307 solutions in D2O at pH 8.3 as a function of temperature (A) and at pH 1.9 (B). SANS curves for 20% T1307 at pH 8.2 (C) and at pH 2.9 (D). Solid lines are fits to the models described in the text.

micelles, whereas at higher temperatures the intensity increases, with the occurrence of a peak reflecting strongly interacting micelles. Fits to a core−shell model combined with a hardsphere potential at 40 and 50 °C yield core radii Rc of 27 and 30 Å, respectively, and a shell of 34 Å, respectively (Table 2). This provides aggregation numbers of 4 and 6, with a shell relatively hydrated compared to the gels of the poloxamine at natural pH. Overall, the effect of lowering the pH concurs with the observations made in dilute solutions (Table 1), in that acidic conditions induce a shrinking of the shell, overall smaller micelles, lower Nagg, and higher degree of hydration. Effect of Cyclodextrins on the Aggregation of T1307. The interaction of cyclodextrins with PEO and PPO and their copolymers has been studied since the work of Harada.6,7,16,22,37,38 Although native CDs have been shown to form insoluble pseudopolyrotaxanes (PPRs) or gels,4,5 soluble PPRs formed with substituted CDs are more likely to find pharmaceutical applications.39 Modified CDs, however, are found generally to have a lower capacity to form PPRslargely because of the reduced hydrogen-bond capacityand tend to disrupt the micellization behavior of PEO−PPO-based block copolymers,21,22 a feature which could be of interest in triggering the release of a cargo from the interior of the

Table 1. Micellar Parameters of 1% T1307 in D2O Obtained from the Analysis of the SANS Data Using a Core−Shell Model with a Hard-Sphere Structure Factora pH

T/°C

Rc/Å

t/Å

ϕ

ρshell × 106/Å−2

Nagg

nsolv/EO

8.3

40 50 40 50

34.9 40.3 25.3 30.8

51.3 50.0 46.4 50.3

0.043 0.040 0.030 0.042

5.97 5.87 6.15 6.06

13 19 4 8

21 16 37 28

1.9

a Rc (core radius), t (shell thickness), ϕ (volume fraction from the hard-sphere potential), ρshell (scattering length density of the hydrophilic corona), Nagg (aggregation number), and nsolv/EO (number of solvent molecules per EO in the shell).

In this situation, the spheres would be in close contact without forming an specific type of paracrystal. It cannot be discarded, however, that at concentrations higher than 20%, more ordered structures may be formed. As shown above (Table 1), acidic pH hinders micelle formation and makes them shrink. This has dramatic consequences as gelation is impeded at pH lower than pKa1. Figure 3D shows the SANS curves from 20% T1307 at pH 2.9. At the lowest temperature (20 °C), the profile could be attributed to either unimers alone or a mixture of unimers and 5649

DOI: 10.1021/acs.langmuir.5b01081 Langmuir 2015, 31, 5645−5655

Article

Langmuir

Table 2. Micellar Parameters of 20% T1307 in D2O Deduced from SANS Data Analysis at Varying Temperatures (°C) and pHa pH

T/°C

Rc/Å

t/Å

ϕ

ρcore × 106/Å−2

ρshell × 106/Å−2

Nagg

nsolv/EO

nsolv/PO

8.2

20 36 50 65 40 50

25.6 53.0 51.3 47.9 27.3 30.0

33.9 22.3 24.7 28.1 33.5 34.4

0.348 0.627 0.628 0.604 0.405 0.426

1.04 2.71 2.45 1.83 3.44 3.44

6.17 6.05 5.94 5.76 6.24 6.14

3 16 17 18 4 6

28 8 8 8 26 20

1 6 4 2 0 0

2.9

a Rc (core radius, Å), t (shell thickness, Å), ϕ (volume fraction from the hard-sphere potential), ρ (scattering length density), Nagg (aggregation number), and nsolv/x (number of solvent molecules per EO or PO units in the shell).

Figure 4. (A) Intensity size distribution of mixtures of 1% T1307 and 5% DIMEB as a function of temperature. (B) Relative excess scattering at 40 °C for mixtures of 1% T1307 with substituted CDs (30:1 CD/poloxamine molar ratio).

water at 35 °C. The CMT is therefore shifted by approximately 15 °C by the presence of DIMEB. Reducing the amount of DIMEB reduces the extent of the CMT shift (from 30 to 35 °C at a 15:1 molar ratio). The other substituted CDs also produce a shift, but it is more limited (about 5 °C at the same molar ratio of 30:1), with the exception of the ionic CD, SBCD, which does not modify the CMT onset (data not shown). A semiquantitative approach to comparing the disruptive effect of different CDs on micellization is achieved by measuring the excess scattering intensity at a temperature where micelles are fully formed (40 °C), relative to the intensity of T1307 at 25 °C, where only unimers are present, at fixed concentrations of poloxamine and CD (Figure 4B). The results show distinctively how DIMEB has by far the most adverse effect against micellization, which seems to be a generic feature of this specific CD,21,28 a result that should be carefully taken into consideration when designing formulations with PPRs based on this modified CD. Nature of CD/Polymer Interactions: NMR Studies. The underlying question here is to ascertain whether the interaction of the modified CDs with the Tetronic induces the formation of a PPR, i.e., whether the CDs thread onto the arms of the poloxamine, or if another type of association is responsible for micellar rupture. There is evidence in the literature pointing in

micelles or the transfer of this cargo from the micellar interior to the cavity of the cyclodextrins at specific target sites.23 Overall, CDs provide a useful tool to manipulate the properties of Tetronic block copolymers, but a detailed understanding of the structures of the assemblies and interactions involved in the complexation is needed to underlie any further formulation. The addition of native cyclodextrins (CDs) to a 1% solution of T1307 indeed induces the formation of white precipitates consisting of the insoluble PPR, a result of the threading of the CDs onto the polymeric chains and as observed with other poloxamines.20 Instead, in the presence of substituted CDs, the solutions remain clear and shift the CMT to higher temperatures, with DIMEB, the twice-methylated CD, being the one which produces the largest shift in the CMT. Figure 4A shows the size distribution as a function of the temperature at a 30:1 molar ratio of DIMEB/T1307 (1% T1307). At 30 °C two overlapping distributions are detected, at ca. 1.3 and 4.0 nm in hydrodynamic radius. On the basis of the size, the smallest one likely corresponds to free DIMEB22 and the other one to the polymer−DIMEB complex, with a size smaller than that of pure unimers in water at the same temperature (5.5 nm). When the temperature is increased, the two modes merge, and at 45 °C, a distinct and slower mode appears at 9 nm, corresponding to the micelles that are exactly the same size as the polymer alone in 5650

DOI: 10.1021/acs.langmuir.5b01081 Langmuir 2015, 31, 5645−5655

Article

Langmuir both directions.7,40 It was recently shown that micellar rupture of the highly hydrophilic Pluronic F127 with DIMEB occurs via very fast kinetics (