Mechanically Enhanced Liquid Interfaces at Human Body

Jan 18, 2016 - Mechanically Enhanced Liquid Interfaces at Human Body Temperature ... Thus, by adjusting the degree of hydrophobicity of metNCC, the ...
3 downloads 3 Views 8MB Size
Article pubs.acs.org/Langmuir

Mechanically Enhanced Liquid Interfaces at Human Body Temperature Using Thermosensitive Methylated Nanocrystalline Cellulose N. Scheuble,*,† T. Geue,‡ S. Kuster,† J. Adamcik,† R. Mezzenga,† E. J. Windhab,† and P. Fischer† †

Institute of Food Nutrition and Health, ETH Zurich, 8092 Zurich, Switzerland Laboratory of Neutron Scattering and Imaging, Paul Scherrer Institut, 5232 Villigen PSI, Switzerland



S Supporting Information *

ABSTRACT: The mechanical performance of materials at oil/ water interfaces after consumption is a key factor affecting hydrophobic drug release. In this study, we methylated the surface of nanocrystalline cellulose (NCC) by mercerization and dimethyl sulfate exposure to produce thermosensitive biopolymers. These methylated NCC (metNCC) were used to investigate interfacial thermogelation at air/water and mediumchain triglyceride (MCT)/water interfaces at body temperature. In contrast to bulk fluid dynamics, elastic layers were formed at room temperature, and elasticity increased significantly at body temperature, which was measured by interfacial shear and dilatational rheology in situ. This unique phenomenon depends on solvent quality, temperature, and polymer concentration at interfaces. Thus, by adjusting the degree of hydrophobicity of metNCC, the interfacial elasticity and thermogelation of the interfaces could be varied. In general, these new materials (metNCC) formed more brittle interfacial layers compared to commercial methylcellulose (MC A15). Thermogelation of methylcellulose promotes attractive intermolecular forces, which were reflected in a change in self-assembly of metNCC at the interface. As a consequence, layer thickness and density increased as a function of temperature. These effects were measured by atomic force microscopy (AFM) images of the displaced interface and confirmed by neutron reflection. The substantial structural and mechanical change of methylcellulose interfaces at body temperature represents a controllable encapsulation parameter allowing optimization of lipid-based drug formulations.



INTRODUCTION Oil-in-water emulsions are extensively used in pharmaceutics for encapsulation and controlled delivery of hydrophobic, active ingredients.1 Stimuli-responsive polymers for emulsification are of special interest, since these materials may undergo phase transitions in the human environment as a result of small changes in pH, temperature, or ionic strength.2,3 One major aspect in pharmaceutical science is to tailor these phase transitions to control drug delivery kinetics.4 The thermoresponsive biopolymer, methylcellulose, is often used as key ingredient to adjust rheological properties of bulk phases and to emulsify active ingredients.5−7 Thermogelation of methylcellulose solutions has been investigated by various researchers, and many structural and mechanistic aspects were analyzed and suggestions for the thermogelation mechanisms concluded.8−14 Gel phase transition temperatures were found between 48 and 60 °C depending on the methoxyl substitution characteristics of the cellulose backbone and the degree of hydration.10,15,16 In bulk systems, these gel phase transition temperatures were engineered toward physiological conditions by incorporating salt and other polymers in methylcellulose suspensions.17,18 However, thermoresponses of methylcellulose oil/water interfaces have never been investigated, although © XXXX American Chemical Society

emulsion stabilization is one major application of the amphiphilic biopolymer. Methylcellulose adsorbs to oil/water interfaces and stabilizes emulsions by lowering the interfacial tension and by providing steric repulsion between droplets.19 Moreover, the rheology of such interfaces can crucially impact on emulsion stabilities. Adsorbed to an interface, rheological behavior may differ from fluid dynamics measured in bulk solutions as dimensions changes from 3D structures to pseudo-2D structures.20 Moreover, various biopolymers adsorb irreversibly to air/ water or oil/water interfaces and therefore accumulate. Collected at the interface, molecules can become highly structured and fluid dynamics of the interfacial layer nonNewtonian.21−23 For these reasons Arboleya and Wilde24 have measured highly viscoelastic layers of methylcellulose at the air/ water interface already at room temperature and at low methylcellulose bulk concentrations. Moreover, mass adsorption of methylcellulose in solution to solid hydrophobic interfaces were found to increase with temperature.25 Thus, Received: November 17, 2015 Revised: January 18, 2016

A

DOI: 10.1021/acs.langmuir.5b04231 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Surface Activity−Wilhelmy Plate Technique. Surface activity of metNCC 1−5, NCC, and commercial methylcellulose (MC A15, Methocel) was measured using the Wilhelmy plate technique. A paper or platinum plate was immersed in a Langmuir trough (KSV Nima, Finland) and filled with clean 10 mM phosphate buffer (pH 7). Solutions containing 2.5 mg/mL material were diluted to 0.1 mg/mL in the filled Langmuir trough, allowing the material to adsorb from the bulk to the air/water interface and form a Langmuir film. With 0.1 mg/mL of metNCC, full coverage of the interface was ensured. MetNCCs were dissolved ultrasonically as published elsewhere27 and filtered with a 0.45 μm PTFE syringe filter (Titan, infochroma AG, CH). NCC was dissolved ultrasonically, whereas MC A15 was hydrated under constant shear within 12 h at 4 °C. Among others, surface activity measurements were used to correlate materials with their substitution characteristics. Since differences in surface activity at air/water interfaces show same tendencies and are more pronounced than at oil/water interfaces, most measurements were performed at the air/water interface. Interfacial Dilatational Rheology. Interfacial dilatational viscoelasticity of metNCC 5 was studied by sinusoidal oscillation of the area of fully developed Langmuir films. Fitting surface pressure responses during constant oscillation (γ = 1%, ω = 10 mHz) allowed calculating interfacial dilatational moduli Ei′ and Ei″ as a function of time.28 For air/water interfaces, a special designed Teflon Langmuir trough (surface area = 11 489 mm2) with two opposite cavities at the corners of the trough allowed to refill the evaporated water over time (1 mL/h at 22 °C, 8 mL/h at 37 °C) using a syringe pump (PHD 2000, Harvard Apparatus, USA) and Milli-Q water. For MCT oil/water interfaces, a stainless steel Langmuir trough was used. Interfacial Shear Rheology. Interfacial shear viscoelasticity was investigated using an interfacial shear rheometer (Physica MCR 501, Anton Paar, Austria) equipped with a biconical geometry.29 The geometry was sinusoidally oscillated between an upper oil phase, medium-chain triglyceride oil (MCT-oil) Delios V (BASF, Germany), and a lower aqueous phase, consisting of solutions described in the previous sections to calculate the interfacial shear moduli G′i and G″i . Bulk rheology is not shown in this article since at this low concentrations of methylcellulose (0.1 mg/mL), no bulk effects could be measured. Neutron Reflection. Neutron reflection of fully developed Langmuir films was investigated using the Swiss Spallation Neutron Source (SINQ) at the Paul Scherrer Institute (Villigen, Switzerland) with the AMOR time-of-flight (TOF) reflectometer. Langmuir films were formed at air/D2O (Armar Chemicals, Switzerland) interfaces as described before. In this setup, flow rates to avoid evaporation were 0.01 mL/h at 22 °C and 4 mL/h at 37 °C. For Langmuir films at the trioctanoin TC8 (Sigma-Aldrich, Switzerland)/D2O interface, an interface was first established with TC8. In total, 130 μL of TC8 was spread (2 mg/mL chloroform) on the air/water surface to obtain a thick TC8/D2O layer. Neutron reflection was recorded at three angles of incidence θ (0.5°, 1.3°, and 2.8°). To obtain broad scattering vector Qz (= 2πθ/λ) range of 0.01−0.15 Å, the neutron wavelength λ of the TOF-reflectometer AMOR was exploited from 3.5 to 12 Å. Morphological and compositional information on interfaces was extracted by fitting the obtained neutron reflectivity curves R(q) with the standard Parratt algorithm (Parratt32, version1.6),30,31 since scattering length density (sld) of the material correlates with neutron reflection. Atomic Force Microscopy (AFM). AFM images of metNCC were taken of displaced Langmuir films using a modified Langmuir− Schaefer technique as detailed in earlier work.32−34 At 22 °C, Langmuir films were transferred after approximately 6 h from the air/ water interface to freshly cleaved mica. The mica was horizontally, shortly dipped onto the Langmuir film at the solution surface, followed by Milli-Q water and carefully dried with pressurized air. For 37 °C, Langmuir films were first equilibrated for 4 h at 22 °C and heated for 2 h to 37 °C prior to the deposition. NCC solutions (1 mg/mL) were pipetted on freshly cleaved mica, incubated for 2 min, rinsed by MilliQ water, and dried by pressurized air. AFM images were performed on

for drug formulations, it is necessary to consider not only the thermogelation of bulk systems but also the fluid dynamics of the interface itself. To our knowledge, fluid dynamics of methylcellulose liquid interfaces has not been studied for temperatures other than ambient temperature. Therefore, we design thermoresponsive substrates based on methylated nanocrystalline cellulose (metNCC) to investigate the role of surface activity on thermogelation at the air/water and MCT oil/water interface at body temperature. Nanocrystalline cellulose (NCC), also called cellulose nanocrystals (CNC) or cellulose whiskers, was chosen as base material because of its high purity and elastic modulus (150 GPa), close to the value of a perfect cellulose crystal.26 Furthermore, we report about how engineered surface activity of metNCC can be used to control the mechanical properties of an interface and compare these results to commercial methylcellulose (MC A15). We show that thermogelation influenced the structure of the interface leading to closed compact interfacial structures ideal for encapsulation. This encapsulation parameter is often neglected since formulations are adapted to thermogelation in methylcellulose bulk solutions. To investigate the thermoresponse of methylcellulose, we combined highly sophisticated interfacial techniques such as interfacial dilatation and shear rheology, neutron reflection, and interfacial atomic force microscopy.



MATERIALS AND METHODS

NCC and Preparation of Different Substituted MetNCC. NCC was kindly provided by Celluforce (Canada). It is a white powder and builds stable suspensions in water. According to the manufacturers, it consists of 100% sulfated cellulose nanocrystals with a crystal size ranging from 2 to 10 nm width and 80−150 nm length. Figure 1 shows an atomic force microscopy image of NCC. NCC was

Figure 1. AFM image of NCC from Celluforce. methylated to produce various metNCC, which differ in their surface activity (called metNCC 1−5 in the following). This was controlled over the sodium hydroxide concentration in the reaction mixture. For the methylation, 1 g of NCC was mercerized with 20 mL of aqueous NaOH solutions (10% w/w for metNCC 1, 20% w/w for metNCC 2, 30% w/w for metNCC 3, 40% w/w for metNCC 4, and 50% w/w for metNCC 5) for 1 h before 3 mL of the methylation agent DMS (dimethyl sulfate, Merck Schuchardt OHG, Germany) was added to this dispersion. For metNCC 1−4 the slurry was methylated in a glass reaction vessel for 3 h at 50 °C while stirring vigorously. For metNCC 5 the excess of aqueous NaOH was removed by centrifugation (10 min, 21000g) prior to the addition of DMS to further increase the degree of substitution. After methylation, the dispersions were neutralized with an aqueous acetic acid solution (10%, VWR, Switzerland) and filtered. The filtration residue was repeatedly washed with excess of acetone and dried in a high vacuum at 40 °C overnight. B

DOI: 10.1021/acs.langmuir.5b04231 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir a Nanoscope VIII multimode scanning probe microscope (Veeco Instruments, USA) in tapping mode in air.

hydrophobicity, increases with degree of dehydration. Here, surface pressures of around 27 mN/m were reached with metNCC 3, where after the increase in surface pressure leveled off. Sarkar35 found for commercial methylcellulose that surface activity of methylcellulose depends mainly on the amount of methoxyl groups but as well on their distribution. More precisely, they stated that more nonuniform distributed methoxyl regions resulted in higher surface pressures. Hence, high surface activities of metNCC 3−5 might originate from a superposition of both the methylation degree and the distribution of the methoxyl groups at the NCC surface. Surface tension of methylcellulose is independent of molar mass and explicitly depends on the substitution characteristics.36 Since methylcellulose is a flexible polymer and NCC a stiff particle, which most likely is only methylated at the surface, we might not directly correlate surface activity with its substitution degree. Nevertheless, it might give an estimation of the apparent substitution degree of the available surface of metNCC particles at the air/water interface. The commercial MC A15 with a substitution degree of 1.6 was taken as a reference; its surface pressure was measured at around 22 mN/ m, which agrees with previous data.24 Thus, with our methylation procedure we obtained lower (metNCC 1), comparable (metNCC 2), and higher (metNCC 3−5) apparent substitution degrees than with MC A15. Thermoresponses of metNCC 1−5 and MC A15 adsorbed to MCT oil/water interfaces were investigated using interfacial shear rheology. The biconical tool was oscillated constantly at the MCT oil/water interface while measuring the interfacial moduli Gi′ and Gi″ as a function of time. After interfacial equilibration, the whole setup was heated to 37 °C. Figure 3A illustrates the experimental procedure for MC A15. It shows how MC A15 forms an elastic interface at 22 °C and how the interface responds to the heat treatment. The interfacial moduli increase more than 10 times when temperature was raised to 37 °C. In bulk, sol−gel transitions of typically 2 wt % MC A15 solutions were measured at more elevated temperature of 48 °C. In bulk rheological measurements, two steps during thermogelation of methylcellulose solutions were defined. Above 30 up to 50 °C methylcellulose solutions started to gel (G′ > G″), whereas between 48 and 60 °C instantaneous and fast thermogelation kinetics were observed (G′ ≫ G″).37−39 Here, the interfacial layer already gelled at 22 °C (G′i ≫ G″i ) and significantly increased at body temperature ′ °C ≫ Gi,22 ′ °C) even though the bulk concentration is very (Gi,37 low (0.01 wt %). Since gelation of methylcellulose depends on the hydration degree and thus the concentration, we conclude that, adsorbed to an interface, methylcellulose accumulates and additionally loses a degree of freedom, causing gelation at low temperatures and an early, significant thermoresponse. Figure 3B summarizes the values for Gi′ and Gi″ from metNCC 1−5 and MC A15 before and after thermogelation. The thermoresponse is significant for all materials. With respect to the initial viscoelasticity at 22 °C, the thermoresponse was less pronounced for metNCC (Gi,37 ′ °C = 3−5Gi,20 ′ °C) than for MC A15 (G′i,37 °C = 11G′i,20 °C), which might be explained by the higher flexibility of the soluble MC A15. The interfacial viscoelasticity strongly correlated with the surface activity of methylcellulose. Hirrien et al.10 highlighted that gelation of methylcellulose solutions originates exclusively from hydrophobic interactions between the methoxyl groups. Therefore, thermal gelation of metNCC and methylcellulose might underlay similar basic principles. Less surface active material



RESULTS AND DISCUSSION Impact of Surface Activity on Interfacial Thermoresponse of Methylcellulose. Surface activity of pure NCC, metNCC 1−5, and MC A15 was studied by measuring surface pressures Π during material adsorption from the bulk to the air/water interface by the Wilhelmy plate technique. Figure 2B

Figure 2. (A) Equilibrated surface pressures Πeq of MC A15, NCC, and metNCC 1−5 at the air/water interface. Lines are to guide the eye. (B) Adsorption of metNCC 5 to the air/water interface as a function of time. Surface pressures were measured with the Wilhelmy plate technique at 22 °C and bulk solutions of 0.1 mg/mL. (C) Schematic diagram of methylation of NCC the correlation to the surface activity.

shows an example of such an adsorption curve, where the equilibrated surface pressure Πeq is extracted and plotted as a function of material in Figure 2A. NCC showed no increase in surface pressure, thus no surface activity, demonstrating the high purity of the material. In contrast, all metNCC types were surface active, whereas surface pressures of metNCC increased with the degree of dehydration during the methylation process. The schematic in Figure 2C illustrates the change in surface structure of NCC during the methylation process. The surface of NCC erodes and deprotonates during mercerization, giving the opportunity to partially replace the functional hydroxyl groups present at the NCC surface by methoxyl groups making it surface active. The eroded products are washed away during our procedure. This is evidenced by AFM imaging discussed later in the article. The methylated surface of NCC might be seen as a methylcellulose gel layer. Higher NaOH concentrations during methylation led to more deprotonated NCC. Hence, more hydroxyl residues are replaced by methoxyl groups at the NCC surface and thus surface activity, i.e., C

DOI: 10.1021/acs.langmuir.5b04231 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 4. (A) Strain sweep of metNCC 3 adsorbed to the MCT oil/ water interface at constant frequency 1 s−1. The box marks the linear viscoelastic region LVR. (B) Critical strain before and after thermogelation as a function of different materials. Bulk phase concentrations were in all experiments 0.1 mg/mL. Lines are to guide the eye.

Figure 3. (A) Gi′ (closed symbols) and Gi″ (open symbols) as a function of time for MC A15. (B) Gi′ and Gi″ as a function of methylcellulose type before and after thermogelation. Constant interfacial shear deformation with γ = 0.3% and ω = 1 s−1 was applied in all experiments. Bulk phase concentrations were in all experiments 0.1 mg/mL. Lines are to guide the eye.

where after the interfacial layer starts breaking. At the crossover of G′i and G″i the layer finally breaks. The critical strain was taken as a measure of brittleness and plotted as a function of material in Figure 4B before and after the thermoresponse. The critical strain decreases with increasing layer strength of metNCC. Thus, metNCC interfaces with a higher viscoelasticity representing stronger hydrophobic interactions were more brittle than interfaces with lower viscoelasticity. More precisely, metNCC 1 with the lowest surface activity and interfacial viscoelasticity was the least brittle layer and metNCC 5 the most brittle one. Increased viscoelasticity due to thermoresponses resulted in more brittle layers. Conversely, the interfaces formed by the flexible polymer, MC A15, were more stable against deformation after heating. Thus, brittleness of methylcellulose interfaces does not only depend on the surface activity and the interfacial viscoelasticity, but rather on the type of methylcellulose, i.e. the raw material and the methylation process. Nanocrystalline cellulose is a stiff crystalline material resulting in methylcellulose interfacial layers, which are less flexible than commercial methylcellulose. Impact of Interfacial Thermogelation on MetNCC Structure and Interfacial Dilatational Rheology at Liquid Interfaces. Morphology and dilatational viscoelasticity of metNCC 5 interfaces (the methylcellulose with highest surface activity and interfacial shear elasticity) were investigated combining neutron reflection, AFM imaging, and interfacial oscillatory experiments at room and body temperature. The areas of equilibrated Langmuir films were constantly sinus-

(such as MC A15, metNCC 1, and metNCC 2) resulted in lower interfacial moduli, since less hydrophobic intermolecular forces are present. After a critical surface activity (for metNCC 3−5), hydrophobic interactions strengthen, resulting in more than 10-fold higher interfacial moduli compared to less surface active material. Several authors10,15,16 concluded that crosslinking zones of methylcellulose consist of pseudocrystalline sequences of highly substituted units (trimethylglucose units), which were determined to be compulsory for gelation. Hence, metNCC 5 showing the highest surface activity resulted in high interfacial moduli indicating not only a high apparent substitution degree of the NCC surface but also a nonuniform distribution of methoxyl regions. To conclude, the interfacial shear viscoelasticity of metNCC depends on its surface activity whereas thermoresponses were proportional to the initial viscoelasticity at 22 °C. Lipase activity depends on the ability of lipases to adsorb to oil/water interphases. Generating a physical barrier for lipase adsorption influences lipase activity and thus lipid/drug release. Tuning this physical barrier by combining surface activity and thermal gelation might allow to control lipid/drug release. Brittleness of the discussed interfacial layers was measured performing strain sweeps at constant frequency before and after the thermoresponse. Figure 4A shows an example of the strain sweep of metNCC 3. The linear viscoelastic region (LVR), where structure remains during deformation, is marked as a box. We defined the border of the LVR as the critical strain, D

DOI: 10.1021/acs.langmuir.5b04231 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

methylcellulose polymers, which might have been produced during our methylation procedure, were detected. Soluble methylcellulose polymers would adsorb fastest (first) to the air/ water interface and appear as background layer on the AFM image. Usually pure mica was detected in the background. Especially the AFM images at 22 °C reveal that the needle shape of nanocrystalline cellulose remained during our harsh treatments. The metNCC 5 needles show a thickness of approximately 3 nm and a length between 50 and 120 nm. They aligned at the interface in a first thin homogeneous layer. Figure 6C indicate that the crystals align rather isotropically. Several spots show multiple layers and larger agglomerates (Figure 6B), which are inhomogeneously distributed. The agglomerates might have existed already before adsorption to the interface. We assume that a layer formed at 22 °C rather diffuses into the bulk phase, which challenges the displacement of the layer and might result in a layer with more height variations. The images at 37 °C show a completely different structure. They indicate a homogeneously, densely packed layer of about 15 nm with lower height variations. Similar effects have been seen by Bodvik et al.44 studying AFM images of methylcellulose solutions adsorbing to silanized silica at different temperatures. They measured an increase of mass adsorption of more than 60%. The structures within the metNCC layer suggest isotropic agglomerates of metNCC 5, which protrude out of the layer and appear as spherical-like objects in the AFM images. During gelation of methylcellulose, the entropy of the overall methylcellulose−water system increases due to an increase in the disorder of the methylcellulose system.12 However, a closer look at our AFM images indicates that a disordered structure aligns to a preformed structure at the interface, which must have been formed at 22 °C. This becomes more evident by looking at sharp edges in the AFM images, which originated from the sample preparation due to the earlier discussed brittleness of this layer. To get more insight into the morphological change between 22 and 37 °C, neutron reflections at air/D2O interfaces were measured (Figure 7A). The reflectivity shift upon heating indicates tremendous interfacial structural changes. Models for the reflection curves were evolved using Parratt30,31 and the knowledge gained from the AFM images. The models are summarized as scattering length density (sld) profiles in Figure 7B and illustrated in Figure 7C. The parameters used for the fits are given in the Supporting Information. Best fits obtained for the layer formed at 22 °C represent a thin layer directly at the interface, which diffuses into the bulk phase. Therefore, sld’s of this diffusive sublayer approach sld’s of deuterium and finally disappear in the bulk phase. In our previous publication27 we misinterpreted this diffusive layer as a rough layer formed from methylcellulose particles protruding from the interface into the bulk phase. The response of the layer to heating resulted in intermolecular attraction and therefore in a solvent exclusion leading to a thick, densely packed layer with a scattering length density dominated from metNCC 5. The best model reveals that the transition from the interface to the air is rougher at 37 °C, whereas the transition to the bulk is less rough and thus less diffusive. Neutron reflection of the system recooled to 22 °C reveals that the thermal induced structural change partially remains. Indeed, the best fit represents a densely packed multilayer of metNCC 5, but the transition from the interface into the bulk phase is more diffusive than at 37 °C. This indicates that close to the air/water interface, where metNCC

oidally oscillated while heating the setup. Figures 5A and 5B show the oscillatory response of the surface pressure Π to

Figure 5. Sinusoidal oscillation of the trough area (dotted line) and the surface pressure Π (continuous line) at (A) room temperature and (B) body temperature. (C) Interfacial dilatational moduli E′i and E″i as a function of time while heating to body temperature. Constant area deformation with γ = 1% and ω = 0.01 s−1 were applied. Bulk phase concentrations of 0.1 mg/mL were used.

defined interfacial area deformations at 22 and 37 °C, respectively. At 22 °C, Π oscillated around 29 mN/m, whereas at 37 °C Π oscillated stronger around a value of 43 mN/m. This drastic increase in Π originates partly from the temperature increase itself (3 mN/m) but must incorporate morphological changes of the interface, which we discuss later on. From the oscillatory experiments, interfacial dilatational moduli E′i and E″i were extracted and plotted as a function of time in Figure 5C. An elastic interface is formed at 22 °C and is comparable to values found by other authors for commercial methylcellulose.24 As soon as the temperature was raised, the interfacial dilatational elasticity started increasing and leveled off after 30 minEi′ almost doubled whereas Ei″ increased slightly. This large increase in interfacial viscoelasticity with temperature is caused by a decrease in solvent quality, which might lead to multilayer formation. Comparable effects have been reported for other thermosensitive polymers using similar techniques.40−43 One main difference to our work is that methylcellulose phase transitions in bulk occur at higher temperatures and higher bulk concentrations. The interfacial dilatational viscoelasticity remain after recooling the system to 22 °C, indicating an irreversible gelling of the interfacial layer. However, investigating the morphological changes of thermoresponsive materials at interfaces is crucial to draw conclusions. Hence, we displaced equilibrated Langmuir adsorption layers directly from the air/water interface at room and body temperature and took AFM images. For both individual layers, different spots on the same mica are depicted in Figure 6. The morphology change of the layers exposed to different temperatures is drastic. In all AFM images, no soluble E

DOI: 10.1021/acs.langmuir.5b04231 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 6. AFM images of displaced Langmuir films formed at the air/water interface at (A−C) room temperature and (D−F) body temperature. (A−C) Displacement of a/w interface after metNCC adsorption for almost 6 h at 22 °C on one mica. (D−F) Displacement of a/w interface after metNCC adsorption for 4 h at 22 °C and subsequent heating to 37 °C almost for 2 h on one mica. Bulk phase concentrations of 0.1 mg/mL were used. AFM was performed in tapping mode in air.

To assess more in detail the potential role of the hydrophobic phase during this thermal induced morphological change, we benchmarked the above results with a control study made at an oil/water interface. Hence, morphology of metNCC 5 at the trioctanoin (TC8)/D2O interface was investigated by neutron reflection at room and body temperature. The neutron reflections are shown in Figure 8A and the sld profiles of the corresponding fits in Figure 8B. A striking shift in neutron reflection with temperature was measured, similar to the shift obtained at the air/D2O interface. Reference measurements of TC8/D2O interfaces revealed that TC8 acted as a cap, since the best fits were obtained by shifting the zero air reflection into the direction of trioctanoin. The sld profiles of the best fits obtained for metNCC 5 at the TC8/D2O interface are similar to the ones evolved for air/D2O. They mainly differ in the

mainly interacts with itself or air, the thermoresponse is irreversible due to an increased surface anchoring energy after gelation. Adsorption hysteresis upon heating and recooling was found for methylcellulose adsorbed to solid hydrophobic surfaces.44 However, for solid hydrophobic surfaces the surface anchoring energy might be lower, since the polymers/particles cannot penetrate into the hydrophobic phase. To summarize, metNCC 5 forms a thin anisotropic, viscoelastic layer with diffusive sublayers at 22 °C. After heating to 37 °C, it is transformed into a compact, highly elastic layer forming a homogeneous membrane with isotropic subunits, which follow a prealignement originating from the interface at 22 °C. This structuring and gelling are only partially reversible upon recooling. F

DOI: 10.1021/acs.langmuir.5b04231 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 8. (A) Neutron reflectivity of fully developed Langmuir adsorption layers of metNCC 5 at 22 and 37 °C at trioctanoin/ deuterium interfaces with its corresponding fits using Parratt. Bulk phase concentrations of 0.1 mg/mL were used. (B) Scattering length density (sld) as a function of layer depth. (C) Interfacial dilatational moduli E′i and E″i of equilibrated Langmuir films of metNCC 5 at the MCT oil/water interface as a function of time while heating to body temperature (area deformation: γ = 1% and ω = 0.01 s−1; c = 0.18 mg/ mL). Surface pressures Π of the equilibrated Langmuir films at the MCT oil/water interface for 22 and 37 °C are embedded.

Figure 7. (A) Neutron reflectivity of equilibrated Langmuir films of metNCC 5 at 22 and 37 °C at air/deuterium interfaces with its corresponding fits using Parratt. Bulk phase concentrations of 0.1 mg/ mL were used. (B) Scattering length density (sld) as a function of layer depth. (C) Illustration of the morphological change of metNCC before and after its thermoresponse.

upon area oscillations, indicating rearrangements in the interface. Upon heating, E′i increases stronger and starts leveling off after less than 1 h. The increase in hydrophobic forces causes additional adsorption and compaction of the metNCC layer, leading to interfacial dilatational elasticities up to 4 times higher than at room temperature. This phenomenon is comparable to thermogelation at air/water interfaces. Therefore, the thermal induced morphology and surface rheology change originate mainly from the diffusive layer in the aqueous bulk phase, whereas the hydrophobic phase plays a subordinate role.

transition region of metNCC 5 and the hydrophobic TC8 phase. Especially at body temperature, the model suggests that metNCC 5 protrudes more strongly into TC8 than into air and thus interacts more strongly with the hydrophobic phase. How this interaction influences the interfacial dilatational rheology is shown in Figure 8C. It shows E′i and E″i of metNCC 5 at the MCT oil/water interface as a function of time upon heating. In general, the interfacial dilatational viscoelasticity and the surface pressures are lower than at air/water interfaces. Lower E′ and E″ may originate from metNCC 5 interactions with MCT oil, and therefore intermolecular forces might be less pronounced. Moreover, initial interfacial tensions of MCT oil/water interfaces (24 mN/m) are lower than those of air/water interfaces (72 mN/m), which might contribute to the lower E′ at MCT oil/water interfaces, since the calculations for E′ and E″ underlay absolute values. In addition, the evaporating flux acting at the air/water but not at the oil/water interface might cause compaction and accumulation of particles leading to higher E′.45 All these effects may superpose, and therefore an absolute explanation for the driving force, why E′ is lower at MCT oil/water interfaces than at air/water interfaces, is not possible. However, E′ and E″ of metNCC at MCT oil/water interfaces are comparable to other data from commercial methylcellulose.46 At room temperature, Ei′ and Ei″ increase



CONCLUSION Controlling the mechanical properties of stimulus-responsive interfaces is crucial for drug formulations. In this particular study, we produced thermoresponsive materials based on methylated nanocrystalline cellulose to investigate the role of surface activity on the mechanical response of liquid interfaces at body temperature. We found that increasing the temperature to 37 °C causes a strong increase in interfacial viscoelasticity representing a significant encapsulating parameter, which depended on the surface activity. The mechanical response strongly correlated to a thermal induced morphological change. The interfacial layer concentrated and compacted during heating, what was evidenced by AFM imaging and neutron reflection. G

DOI: 10.1021/acs.langmuir.5b04231 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

(11) Sarkar, N.; Walker, L. Hydration−Dehydration Properties of Methylcellulose and Hydroxypropylmethylcellulose. Carbohydr. Polym. 1995, 27, 177−185. (12) Lam, Y.; Joshi, S.; Tan, B. Thermodynamic Characteristics of Gelation for Methyl-cellulose Hydrogels. J. Therm. Anal. Calorim. 2007, 87, 475−482. (13) Hirrien, M.; Desbrières, J.; Rinaudo, M. Physical Properties of Methylcelluloses in Relation with the Conditions for Cellulose Modification. Carbohydr. Polym. 1996, 31, 243−252. (14) Arisz, P.; Kauw, H.; Boon, J. Substituent Distribution Along the Cellulose Backbone in O-methylcelluloses using GC and FAB-MS for Monomer and Oligomer Analysis. Carbohydr. Res. 1995, 271, 1−14. (15) Kato, T.; Yokoyama, M.; Takahashi, A. Melting Temperatures of Thermally Reversible Gels IV. Methylcellulose-water Gels. Colloid Polym. Sci. 1978, 256, 15−21. (16) Khomutov, L.; Ryskina, I.; Panina, N.; Dubina, L. G.; Timofeeva, G. N. Structural Changes during Gelation of Aqueous Solutions of Methylcellulose. Polym. Sci. 1993, 35, A276−A279. (17) Xu, Y.; Li, L.; Zheng, P.; Lam, Y. C.; Hu, X. Controllable Gelation of Methylcellulose by a Salt Mixture. Langmuir 2004, 20, 6134−6138. (18) Mayol, L.; De Stefano, D.; De Falco, F.; Carnuccio, R.; Maiuri, M. C.; De Rosa, G. Effect of Hyaluronic Acid on the Thermogelation and Biocompatibility of its Blends with Methyl Cellulose. Carbohydr. Polym. 2014, 112, 480−485. (19) Dickinson, E. Hydrocolloids as Emulsifiers and Emulsion Stabilizers. Food Hydrocolloids 2009, 23, 1473−1482. (20) Edwards, D. A.; Brenner, H.; Wasan, D. T. Interfacial Transport Processes and Rheology; Butterworth-Heinemann: Stonheam, MA, 1991. (21) Fuller, G. G. In Rheology Reviews; Binding, D. M., Walters, K., Eds.; The British Society of Rheology: Aberystwyth, UK, 2003; pp 77−124. (22) Sagis, L. M.; Fischer, P. Nonlinear Rheology of Complex FluidFluid Interfaces. Curr. Opin. Colloid Interface Sci. 2014, 19, 520−529. (23) Murray, B. S. Rheological Properties of Protein Films. Curr. Opin. Colloid Interface Sci. 2011, 16, 27−35. (24) Arboleya, J.-C.; Wilde, P. J. Competitive Adsorption of Mixed Anionic Polysaccharides at the Surfaces of Protein-Coated Lipid Droplets. Food Hydrocolloids 2005, 19, 485−491. (25) Bodvik, R.; Macakova, L.; Karlson, L.; Thormann, E.; Claesson, P. Temperature-Dependent Competition between Adsorption and Aggregation of a Cellulose Ether−Simultaneous Use of Optical and Acoustical Techniques for Investigating Surface Properties. Langmuir 2012, 28, 9515−9525. (26) Miao, C.; Hamad, W. Cellulose Reinforced Polymer Composites and Nanocomposites: a Critical Review. Cellulose 2013, 20, 2221−2262. (27) Scheuble, N.; Geue, T.; Windhab, E. J.; Fischer, P. Tailored Interfacial Rheology for Gastric Stable Adsorption Layers. Biomacromolecules 2014, 15, 3139−3145. (28) Loglio, G.; Tesei, U.; Clini, R. Measurement of Interfacial Dilational Properties: A Software-driven Apparatus. Rev. Sci. Instrum. 1988, 59, 2045−2050. (29) Erni, P.; Fischer, P.; Windhab, E.; Kusnezov, V.; Stettin, H.; Läuger, J. Stress- and Strain-controlled Measurements of Interfacial Shear Viscosity and Viscoelasticity at Liquid/Liquid and Gas/Liquid Interfaces. Rev. Sci. Instrum. 2003, 74, 4916−4924. (30) Parratt, L. G. Surface Studies of Solids by Total Reflection of XRays. Phys. Rev. 1954, 95, 359−369. (31) Braun, C. Parratt32: The Manual; HMI: Berlin, 1999. (32) Jördens, S.; Isa, L.; Usov, I.; Mezzenga, R. Non-equilibrium Nature of Two-dimensional Isotropic and Nematic Coexistence in Amyloid Fibrils at Liquid Interfaces. Nat. Commun. 2013, 4, 1917. (33) Jördens, S.; Rühs, P. A.; Sieber, C.; Isa, L.; Fischer, P.; Mezzenga, R. Bridging the Gap between the Nanostructural Organization and Macroscopic Interfacial Rheology of Amyloid Fibrils at Liquid Interfaces. Langmuir 2014, 30, 10090−10097.

In much higher concentrated methylcellulose bulk systems, thermoresponses became only significant above 40 °C. Hence, thermoresponsive materials at interfaces can react substantially differently than in the bulk. At interfaces, surface-active material adsorbs and concentrates, leading to lower apparent degrees of hydration at the interface. Intermolecular forces increase with surface activity; thus, solvent exclusion and therefore thermoresponses got more significant. These results enhance our understanding of the physical principles of interfaces and the role of surface activity on thermoresponsive materials. Furthermore, they will help for the improved design of engineered dynamic surfaces for controlled mechanical properties of interfaces. Our ongoing research focuses on how such thermoresponsive material can be used to control lipid digestion and thus hydrophobic drug release.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.5b04231. Tables 1−4 (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail [email protected]; Ph +41 (0)44 632 33 04 (N.S.). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors thank Celluforce for providing nanocrystalline cellulose. This work is based on experiments performed at the Swiss Spallation neutron source SiNQ, Paul Scherrer Institut, Villigen, Switzerland. The Swiss National Foundation (SNF) is acknowledged for funding Project No. 2000-21137941.



REFERENCES

(1) Lawrence, M.; Rees, G. D. Microemulsion-based Media as Novel Drug Delivery Systems. Adv. Drug Delivery Rev. 2000, 45, 89−121. (2) Gil, E.; Hudson, S. Stimuli-reponsive Polymers and their Bioconjugates. Prog. Polym. Sci. 2004, 29, 1173−1222. (3) Mano, J. Stimuli-Responsive Polymeric Systems for Biomedical Applications. Adv. Eng. Mater. 2008, 10, 515−527. (4) Baumgartner, S.; Planinšek, O.; Srčič, S.; Kristl, J. Analysis of Surface Properties of Cellulose Ethers and Drug Release from their Matrix Tablets. Eur. J. Pharm. Sci. 2006, 27, 375−383. (5) Kumar, V.; Banker, G. Chemically-modified Celldlosic Polymers. Drug Dev. Ind. Pharm. 1993, 19, 1−31. (6) Douaire, M.; Norton, I. T. Designer Colloids in Structured Food for the Future. J. Sci. Food Agric. 2013, 93, 3147−3154. (7) Ford, J. L. Thermal Analysis of Hydroxypropylmethylcellulose and Methylcellulose: Powders, Gels and Matrix Tablets. Int. J. Pharm. 1999, 179, 209−228. (8) Heymann, E. Studies on Sol-gel Transformations. I. The Inverse Sol-gel Transformation of Methylcellulose in Water. Trans. Faraday Soc. 1935, 31, 846−864. (9) Sarkar, N. Thermal Gelation Properties of Methyl and Hydroxypropyl Methylcellulose. J. Appl. Polym. Sci. 1979, 24, 1073− 1087. (10) Hirrien, M.; Chevillard, C.; Desbrières, J.; Axelos, M.; Rinaudo, M. Thermogelation of Methylcelluloses: New Evidence for Understanding the Gelation Mechanism. Polymer 1998, 39, 6251−6259. H

DOI: 10.1021/acs.langmuir.5b04231 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir (34) Jördens, S.; Riley, E. E.; Usov, I.; Isa, L.; Olmsted, P. D.; Mezzenga, R. Adsorption at Liquid Interfaces Induces Amyloid Fibril Bending and Ring Formation. ACS Nano 2014, 8, 11071−11079. (35) Sarkar, N. Structural Interpretation of the Interfacial Properties of Aqueous Solutions of Methylcellulose and Hydroxypropyl Methylcellulose. Polymer 1984, 25, 481−486. (36) Nasatto, P.; Pignon, F.; Silveira, J.; Duarte, M.; Noseda, M.; Rinaudo, M. Interfacial Properties of Methylcelluloses: The Influence of Molar Mass. Polymers 2014, 6, 2961−2973. (37) Sarkar, N. Kinetics of Thermal Gelation of Methylcellulose and Hydroxypropylmethylcellulose in Aqueous Solutions. Carbohydr. Polym. 1995, 26, 195−203. (38) Li, L.; Shan, H.; Yue, C. Y.; Lam, Y. C.; Tam, K. C.; Hu, X. Thermally Induced Association and Dissociation of Methylcellulose in Aqueous Solutions. Langmuir 2002, 18, 7291−7298. (39) Nasatto, P.; Pignon, F.; Silveira, J.; Duarte, M.; Noseda, M.; Rinaudo, M. Influence of Molar Mass and Concentration on the Thermogelation of Methylcelluloses. Int. J. Polym. Anal. Charact. 2015, 20, 110−118. (40) Guillermic, R.-M.; Saint-Jalmes, A. Dynamics of Poly-nipam Chains in Competition with Surfactants at Liquid Interfaces: from Thermoresponsive Interfacial Rheology to Foam. Soft Matter 2013, 9, 1344−1353. (41) Volden, S.; Genzer, J.; Zhu, K.; Ese, M.-H. G.; Nyström, B.; Glomm, W. R. Charge- and Temperature-dependent Interactions between Anionic Poly(N-isopropylacrylamide) Polymers in Solution and a Cationic Surfactant at the Water/Air Interface. Soft Matter 2011, 7, 8498−8507. (42) Monteux, C.; Mangeret, R.; Laibe, G.; Freyssingeas, E.; Bergeron, V.; Fuller, G. Shear Surface Rheology of Poly(Nisopropylacrylamide) Adsorbed Layers at the Air/Water Interface. Macromolecules 2006, 39, 3408−3414. (43) Fujie, T.; Park, J. Y.; Murata, A.; Estillore, N. C.; Tria, M. C. R.; Takeoka, S.; Advincula, R. C. Hydrodynamic Transformation of a Freestanding Polymer Nanosheet Induced by a Thermoresponsive Surface. ACS Appl. Mater. Interfaces 2009, 1, 1404−1413. (44) Bodvik, R.; Thormann, E.; Karlson, L.; Claesson, P. M. Temperature-dependent Adsorption of Cellulose Ethers on Silica and Hydrophobized Silica Immersed in Aqueous Polymer Solution. RSC Adv. 2011, 1, 305−314. (45) Kabalnov, A.; Wennerström, H. Diffusion in evaporating solutions. Soft Matter 2009, 5, 4712−4718. (46) Torcello-Gómez, A.; Foster, T. J. Interactions between Cellulose Ethers and a Bile Salt in the Control of Lipid Digestion of Lipid-based Systems. Carbohydr. Polym. 2014, 113, 53−61.

I

DOI: 10.1021/acs.langmuir.5b04231 Langmuir XXXX, XXX, XXX−XXX