Polymer Nanocomposite Hydrogels Exhibiting Both Dynamic

May 21, 2013 - Shear-thinning gel-like materials were rarely reported; these materials can flow under high shear stress and recover their solid struct...
0 downloads 13 Views 3MB Size
Download PDF
Article pubs.acs.org/Langmuir

Polymer Nanocomposite Hydrogels Exhibiting Both Dynamic Restructuring and Unusual Adhesive Properties Mian Wang,† Du Yuan,‡ Xiaoshan Fan,‡ Nanda Gopal Sahoo,† and Chaobin He*,†,‡ †

Institute of Materials Research and Engineering, 3 Research Link, Singapore 117602 Department of Materials Science and Engineering, National University of Singapore, 9 Engineering Drive 1, 117576, Singapore

‡

S Supporting Information *

ABSTRACT: Polymer nanocomposite (NC) hydrogels exhibiting both dynamic restructuring and unusual adhesive properties in wet and dry states have been prepared in an efficient and straightforward way via free radical polymerization of poly(ethylene glycol) methyl ether acrylate (PEG) in the presence of silane-modified sodium montmorillonite (NaMMT). The dynamic restructuring of the NC gel has been demonstrated by almost instant recovery of mechanical properties, such as storage modulus, loss modulus, and damping tan δ (at 0.025 strain) by 60−110% after being stressed to the point of gel failure. Furthermore, the dry NC gel showed exceptional thermal and mechanical stability during a heating and cooling cycle between 25 and 110 °C, with only slightly decreases followed by at least 30% increases in both moduli, while tan δ remained nearly unchanged. The NC gel in dry state could repeatedly adhere to various surfaces such as steel, glass, plastic, etc., and detach from the surface without being broken and leaving little contamination behind. This unique adhesive characteristic was characterized by high storage modulus, loss modulus (kPa), and tan δ (>0.6) corresponding to high cohesive, adhesive, and tacking properties of pressure-sensitive adhesives (PSAs). Finally, a reversible network structure formed by PEO interpenetrating within 3-dimentional (3-D) silica network was proposed to be responsible for the dynamic restructuring and the unique adhesive behaviors observed in the NC gel, and the 3-D network structure was investigated by XRD, FTIR, and DSC measurements. For this 3-D network structure, we suggest that the flexibility of PEO could allow PEO side chains to contact with various surfaces by either PEO segments or methoxy end groups via weak physical interactions, such as van der Waals interactions or hydrogen bonding, whereas the reversible network structure contributes to the recovery of strength and shape after the gel failure.

1. INTRODUCTION Soft gel-like materials in nature show time-dependent mechanical properties resulting from dynamic bonding interactions between biomacromolecules, thus exhibiting a balanced solid- and fluid-like behavior.1 Much attention has been focused on the self-assembly and temperature-, pH-, or ion-sensitive transitions of gel-like materials between liquid and solid states. Shear-thinning gel-like materials were rarely reported; these materials can flow under high shear stress and recover their solid structure and elastic strength after shear, showing great potential in the applications of tissue engineering.2 The advantage of using poly(ethylene glycol) in hydrogels over other synthetic polymers has long been established because of its excellent biocompatibility and antifouling property.3−8 Polymer nanocomposite (NC) hydrogel based on clay has been rapidly developed as self-healing materials.9−13 For example, the introduction of a small amount of polyvalent dendrimer to Laponite aqueous suspension produced NC hydrogels in several minutes, and the reunion of two pieces of the gel could be accomplished at room temperature by pressing them together.12 In another study, polyamide hydrogels fabricated with Laponite as physical cross-linker could also © XXXX American Chemical Society

recombine by just contacting the cut surfaces together at mildly elevated temperature.13 In the former case, the author suggested that the complexation between guanidinium ion groups of the dendrimer and the clay dispersed by sodium polyacrylate was responsible.12 In the latter case, H-bonding among amide groups of the polymer and hydroxyl groups of the clay was thought to be the reason.13 In these two cases, Laponite (XLG, gel forming grade) used is commercially synthetic clay of quite small size (about 25 nm in diameter, 1 nm in thickness) and could form gels at very low concentrations of 2 wt % in the presence of some electrolyte, such as electrolyte in tap water, and polyvalent dendrimer. Compared to Laponite, sodium montmorillonite (NaMMT) is naturally available but rarely used for the synthesis of selfhealing NC hydrogels due to their much bigger sizes (hundreds of nanometers to several micrometers in diameter) and tendency to precipitate in aqueous solutions. Thus, using NaMMT instead of Laponite for polymer/clay hydrogel Received: April 4, 2013 Revised: May 16, 2013

A

dx.doi.org/10.1021/la401269p | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

preparation will be more cost-effective, but prevention of NaMMT precipitation during syntheses is a challenge. Here, we present our study which might be the first report on NaMMT-based polymer NC hydrogels exhibiting both dynamic-restructuring at 25 °C and unusual adhesive properties in a temperature range of 25−110 °C, similar to pressure sensitive adhesives (PSA), which have strong adhesive and cohesive strength (storage and loss moluli of kPa as well as high damping tan δ >0.5). The synthesis of the NC hydrogel was performed by free radical polymerization of poly(ethylene glycol) acrylate in the presence of a small amount of clay (2.8− 5.6 wt %) modified with silane. The dynamic restructuring and adhesive properties were studied with rheological tests, and the probable relationship between the properties and the structure of the gel was investigated by XRD, FTIR, and DSC analyses.

2. EXPERIMENTAL SECTION 2.1. Materials. Triethoxyphenylsilane (TEPS), N,N,N′,N′-tetramethylethylenediamine (TMEDA), and ammonium peroxydisulfate (APS) from Sigma-Aldrich were used as received. Poly(ethylene glycol) methyl ether acrylate (PEG, Mn = 480) and poly(ethylene glycol) diacrylate (PEGDA, Mn = 250) from Sigma-Aldrich were passed through a column of neutral alumina oxide. The pristine clay is the sodium montmorillonite (PGW) from Nanocor Inc., with a cation exchange capacity (CEC) of 145 mequiv/100 g, an aspect ratio of 200−400, a d001 spacing of 1.21 nm, and a specific density of 2.6 g/ cm3, and it was used as received. 2.2. Synthesis of Polymer Nanocopmposite (NC) Hdyrogels. Silylation of clay (Na-MMT) was conducted by dispersing 1.5 g of clay in 100 mL of DI water, followed by sonication for 30 min, and then stirred with a magnetic bar for 1 day. 50 mL of ethanol and 5 mL of toluene were added into the clay aqueous suspension. 0.5 mL of TEPS was added to the suspension, followed by sonication for 1 h, and then the mixture was stirred for 20 h. Finally, the silane modified clay was washed with ethanol and DI water at least three times each and redispersed in DI water for later use. Synthesis of PEG nanocomposite (NC) was described as follows: 0.5 mL of clay suspension was mixed with 0.5 mL of poly(ethylene glycol) methyl ether acrylate (0.545 g, PEG, Mn = 480) in 5 mL of DI water for several minutes in a glass vial under magnetic stirring at room temperature, and then 11 mg of APS (2 wt % of the monomer) was added and stirred until it was dissolved. Polymerization was started by the addition of 21 μL of TMEDA (15.5 mg 3 wt % of the monomer). The viscosity of the system increased quickly in less than 20 min, and the reaction was allowed to proceed at room temperature for 1 day to ensure complete polymerization. Chemical PEG gel was synthesized following the same procedure except that a chemical cross-linker PEGDA instead of NaMMT was used. 2.3. Rheology Measurement. Dynamic rheology was performed using a cone−and-plate configuration (20 mm diameter, 1-degree cone geometry) on a stress-controlled rheometer. The gap distance between the cone and the plate was fixed at 0.5 mm. Stress−amplitude sweeps were performed at a constant oscillation frequency of 1 Hz for the strain range of 0.01 to 100 at 25 °C. Oscillatory frequency sweeps were performed at a controlled oscillatory stress (ranging from 1 to 1000 Pa) determined from the linear viscoelastic region of oscillatory stress sweeps performed on each gel condition. The measurement of yield point were carried out by applying a linearly increasing stress (0.1− 10000 Pa in 3 min) to the NC gels and recording the resulting deformation γ = f(σ). The yield point was calculated as the intersection of the lines extrapolated from the linear parts of the experiment curve with logarithmic distribution (Figure 1B). Temperature sweeps were done in a stress (CS) mode, in which the stress value was selected to be in the linear viscoelastic range of the dry gel, previously assessed by stress sweep experiments. Recovery of the gel postfailure was determined by inducing gel failure by at least 1 min of high-amplitude oscillatory stress (1000−10 000 Pa, 1 Hz) and monitoring G′, G″, and γ restoration in oscillatory time sweeps using small-amplitude

Figure 1. (A) Stress−amplitude sweeps were performed at a constant oscillation frequency of 1 Hz for a strain range of 0.01−100 at 25 °C. (B) At a fixed frequency of 1 Hz, increasing the strain amplitude reveals a yield point (in parts A and B: red color, NC-gel-F1; black color, NC-gel-F2; blue color, NC-gel-F1P; green color, NC-F1T.) (C) Storage modulus (closed blue rectangle), yield point (closed red circle), tan δ (green cross), and critical strain γc (pink star) as functions of water contents for NC-gel-F1. deformation conditions (1−100 Pa, 1 Hz). All experiments were performed on triplicate gel samples. 2.4. FTIR and ATR-FTIR Spectroscopy. Fourier transform infrared spectroscopy was carried out using a PerkinElmer 2000 spectrometer. For FTIR testing, the modified clay was washed with ethanol and water three times each, and the resulting solid was dried at 80 °C for 24 h. Nanocomposite hydrogels were dried at 50 °C overnight. KBr disks were prepared by mixing the test samples with dry KBr. Analyses were performed in absorbance mode in the 400− 4000 cm−1 range, with a resolution of 2 cm−1 and accumulation of 16 scans. Attenuated total reflectance (ATR) FTIR was conducted in B

dx.doi.org/10.1021/la401269p | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 2. Dynamic restructuring polymer nanocomposite (NC) gel at 25 °C and 1 Hz. Failure was induced by large-amplitude oscillatory stress and recovery was determined by monitoring the restoration of storage modulus, loss modulus, and strain in 60 min (A: NC-gel-F1, 1 min under 1000 Pa followed by 60 min under 1 Pa; B: NC-gel-F2, 2 min under 10000 Pa followed by 60 min under 10 Pa; C: NC-gel-F1P, 1 min under 10 000 Pa followed by 60 min under 50 Pa; D: NC-F1T, 1 min under 10 000 Pa followed by 60 min under 100 Pa). G′ and G″ as a function of time are normalized to the prefailure moduli calculated from Figure 1A to facilitate comparison of samples with different moduli. absorbance mode in 4000−2500 cm−1 range, with a resolution of 8 cm−1 and accumulation of 256 scans. 2.5. X-ray Diffraction. XRD patterns were obtained using a Bruker X-ray diffractometer (equipped with a two-dimensional detector) in reflection mode. Tests were carried out using nickelfiltered Cu KR1 radiation (λ = 0.154 18 nm) under a voltage of 40 kV and a current of 40 mA. For nanocomposite hydrogels, a piece of silica (100) wafer was used to support the samples. 2.6. DSC and TGA Measurement. Differential scanning calorimetry (DSC) measurements were performed using a TA Instruments Q100 differential scanning calorimeter from 30 to 200 °C at a scan rate of 10 °C/min under a nitrogen flow rate of 50 mL/ min. After the first heating, the sample was kept at 200 °C for 2 min before being cooled at the same scan rate. For the second heating, the sample was equilibrated at 25 °C for 5 min before being heated to 200 °C at a scan rate of 10 °C/min. The accurate content of clay in the nanocomposite hydrogel was determined by TGA analysis which was conducted on a TA Instruments Q500 thermogravimetric analyzer. The loading and degradation temperature of nanocomposite hydrogels were determined by heating sample to 100 °C at 10 °C/min and keeping it at this temperature for 5 min and then heating it to 800 °C at 10 °C/min under N2 flow of 60 mL/min. The water contents were determined as weight losses at 200 °C by TGA analysis because DSC study shows that all of free and bonded water in the gels could be removed under this temperature. Fresh polymer nanocompsoite gels containing 2.8 and 5.6 wt % clay are denoted as NC-gel-F1 and NCgel-F2, respectively, with polymer content fixed at 10.2 wt % of total NC gel weights. Partially dried NC-gel-F1 containing 17 wt % water is denoted as NC-gel-F1P. Totally dried NC-gel-F1 without water is

denoted as NC-F1T, which actually is a polymer composite according to the definition of hydrogels.

3. RESULTS AND DISCUSSION 3.1. Dynamic Restructuring of Polymer Nanocomposite (NC) Hydrogel in Both Wet and Dry States. In Figure 1, NC-gel-F1 in both wet and dry states and NC-gel-F2 in wet state exhibit a plateau modulus G′ (and G′ > G″) at small strains and begin to flow after a crossover G′ = G″ with strain (or stress) increasing, indicating a typical shear thinning behavior of the NC gels.2 This trend can be simply explained as follows: at the lower shear strains intermolecular and topological entanglements of PEG in 3-D silica structure disrupted by the imposed deformation are replaced by new interactions between different segments of the polymer with clay, with no overall change in the extent of entanglement, and therefore the dynamic mechanical properties were independent of strain. The onset of pronounced shear occurs when the externally imposed strain becomes greater than the rate of formation of entanglement. For comparison purposes, the storage modulus, G′, and loss modulus, G″, at 0.025 strain are denoted as initial storage modulus G′0 and loss modulus G″0. In Figure 1C, G′0 and the yield stress (point) of the gel increases by 2 or 3 orders of magnitude with the evaporation of water, from several Pa for NC-gel-F1 to thousands of Pa for NC-F1T. This could be explained by the increased friction of PEO chain in a more compact network structure of clay after depletion of C

dx.doi.org/10.1021/la401269p | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

In the recovery measurements, stain at the end of the test could be indicative of the ability of the gels to restore its original shape (or release internal strain build up during gel failure) under minor stresses. For the all four gels, the strain decreases with increasing G′0 values of the gels (Figure 1), in agreement with that the degree of shape recovery depends on the ability of materials to store deformation energy. The greater the G′, the higher the degree of recovery of the NC gels. In our study the totally dry NC-F1T having greatest G′0 shows the best shape recovery even under the highest shear stress (100 Pa). Combining all these results for NC-gel-F1 having different water contents and NC-gel-F2, it could be concluded that the clay content is associated with the strength and restructuring (reversible cross-links) ability of the NC gel, whereas water might make PEO chain rigid by forming hydrogen bonding and reduce effective cross-linking strength between PEO and clay as a plasticizer. 3.2. Unusual Adhesive Properties of the Polymer Nanocomposite (NC) Hydrogel. We found that NC-gel-F1 in both fresh and dry NC states could stick to and detach from various smooth surfaces, such as glass, steel, plastics, etc., repeatedly without breaking and leaving little contamination behind as shown in Figure 3. In contrast, NC-gel-F2 showed

water. Remarkably, with the decrease of water contents, the damping tan δ of NC-gel-F1 nearly remains constant, which can be ascribed to the great flexibility of PEO side chain, because its Tg is lower than room temperature. On the contrary, the critical strain at which the storage modulus G′ starts to break down decreases for NC-gel-F1 containing less water, suggesting that the linear range of strain−stress was shorter for the dry NC gel, which might be ascribe to the stronger and more effective filler−filler interactions within the 3-D clay network. These interactions also play a dominant role in decrease in the threshold strain amplitude for the onset of the shear thinning with increasing silicate loadings.14 The storage modulus and yield stress of the NC gel also increase with clay content due to the increased topological interaction between polymer and clay. The recovery of the NC gels in wet and dry state was investigated at 25 °C by subjecting the gels under high amplitude oscillatory stresses (1000−10 000 Pa, 1 Hz) for at least 1 min to gel failure and monitoring both moduli (G′, G″) and shape (γ) restoration in oscillatory time sweeps.1 G′ is related to the amount of stored energy during the deformation. After the load is removed, this energy is completely available, now acting as the driving force for the reformation process which partially or completely compensates the previously obtained deformation of the structure. Materials that are storing the whole deformation energy are showing completely reversible deformation behavior since they recover completely with an unchanged shape after a load cycle.15 In Figure 2A, for NC-gel-F1 with the minimum clay content (2.8 wt %), both storage molulus G′ and loss modulus G″ restore 140% and 260% of its initial values of moduli, respectively, in 60 min after gel failure. However, during the whole observing period, tan δ remains larger than 1, indicating that the gel was transformed from solid to liquid irreversibly under high oscillatory stress on experiment time scale. In contrast, as shown in Figure 2B, for NC-gel-F2 containing 5.8 wt % of clay, G′ and G″ crossovers at 65 s as implied by tan δ = 1, and after that G′ keeps larger than G″ until the former reaches 97% and the latter 92% of their initial values, indicating the recovery of both solid shape and mechanical properties. Furthermore, the final strain at 3600 s for NC-gel-F2 is only 0.033 compared to 0.21 for NC-gel-F1. This suggests that clay as a cross-linker in the NC gels contributes not only to the gel strength but also to the elasticity or structure resilience. Even though fresh NC-gel-F1 with lower clay content does not show recovery of solid structure and properties after gel failure, both partially and totally dried NC-gel-F1 exhibit amazing dynamic restructuring behavior as shown in Figure 2C,D. On one hand, both NC-gel-F1P and NC-F1T restore their solid structure in less than 5 min as indicated by tan δ beginning to become lower than 1 (G′ > G″) from 208 s for the former and 89 s for the latter. Compared to NC-gel-F2, NCgel-F1P and NC-F1T exhibit slower liquid-to-solid recovery, which might be related to the lower clay content and lower cross-linking density, whereas the relatively quicker recovery of NC-F1T compared to NC-gel-F1P could be ascribed to the more compact cross-linking structure which could also account for the higher degree of strength recovery of the former. The totally dried NC-F1T shows no less than 100% restoration of both G0′ and G0″ compared to 61% and 74% for G0′ and G0″, respectively, for NC-gel-F1P. On the other hand, the strain (γ) at 3600 s is decreased from 0.21 to 0.033, 0.015, and 0.004 for NC-gel-F1, NC-gel-F2, NC-gel-F1P, and NC-F1T, respectively.

Figure 3. Adhesive behavior of NC-gel-F1 in fresh and dry states. A: Stretching of NC-gel-F1; B: stretching of NC-gel-F1P; C: NC-gel-F1p sticking to two plastic cubes; D: NC-gel-F1p adhering to steel plane; E: and stretching by external force; F: finally detached from the plane without breaking and leaving no contamination on the steel plane.

very weak tackability to the surfaces used for the adhesion test and broke easily under pressing and stretching. The adhesive behavior of partially and totally dry NC-gel-F1 in one way is similar to pressure sensitive adhesive (PSA) (self-adhering bandage) which is defined as dry adhesive materials possessing a lasting and aggressive tack which enables them to adhere to a wide variety of substrates, upon contact. The performance characteristics of a PSA can be characterized by rheological techniques. For example, oscillatory frequency sweeps are suitable for characterizing the bonding and debonding behavior of a PSA. In the literature, Chu described that the storage modulus at high frequency might be related to peel or quick stick tests, and at low frequencies, it is related to shear resistance of adhesives,16 whereas the tacky behavior can be evaluated and controlled using the loss factor tan δ and a sample shows stickiness only if the tan δ value is in a medium range.17 Frequency sweep measurements show that the fresh and dry gels follow different models (Figure 4). For fresh gels, in the medium frequency range (0.1−60 Hz), storage moduli G′ of both NC-gel-F1 and D

dx.doi.org/10.1021/la401269p | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

PEO interacts with silica spheres, we consider that the different frequency dependence of two moduli for fresh and dry gels are attributed to the change in the configuration of the interpenetrating network formed by clay and PEG with different water contents and under varied oscillatory shears.19 Since dry gels remain as viscoelastic solids under very small oscillatory stresses over the whole frequencies investigated, we further examined the effect of larger stresses on the frequencydependence of G′, G″, and tan δ using completely dry gel NCF1T as a model sample. As shown in Figure 4 C, both G′ and G″ decrease slightly with increasing shear stresses at low frequencies but are nearly the same at higher frequencies, with G′ > G″ remaining unchanged. Since G′ does not exceed G″ by orders of magnitude as would be expected from a true solid, this might be indicative of a “pseudo-solid”-like behavior, which was also seen in PCL/silica polymer nanocomposites.20 The power-law exponent n for G′ changes from 0.35 to 0.32 and to 0.29 as shear stresses decreases, in contrast to 0.43 for NC-gelF1P, indicating a more frequency-dependent elastic modulus of the latter. Notably, damping tan δ of NC-F1T is slightly frequency-dependent, with values around 0.7, suggesting a great tacky nature of the totally dry gel which will be discussed later.16,17 A PSA (press sensitive adhesive) often has high G′ and G″ values (thousand Pa) and tan δ values within a range of 0.4−0.7 in a specific application. The high G′ and G″ usually indicate a high cohesive and adhesive strength of adhesive itself, while a high tan δ value implies good tack to target surfaces.17 As shown in Figure 4, in the whole frequency range, only NC-gelF1P and NC-FT exhibit storage modulus in the range of 200− 2000 and 200−5500 Pa, respectively, with tan δ remaining slightly unchanged around 0.69 and 0.76, respectively, which correspond to both high adhesion and good tackability of the NC gels. Whereas for NC-gel-F2, a relatively low storage modulus together with a low tan δ value is consistent with its ease to break and poor tackability during adhesion test. The difference between NC-gel-F1P and NC-FT and PSAs is that at high frequencies our gels remain in solid state (G′ > G″) with its adhesive properties unaltered, whereas PSAs transform from solid to liquid state. Compared to PSAs, NC-FT not only shows structure stability at high frequencies with good adhesion properties but also exhibits excellent thermal stability without compromising adhesive properties too much. As shown in Figure 5, during heating cycle, both storage modulus and loss modulus of the gel decrease with temperature initially and then increase and recover nearly 78% and 51% of its original values, respectively, at the highest temperature investigated. Amazingly, when the gel is cooled to room temperature, both G′ and G″ increase up to 135% and 170% of its initial values, indicating a stronger adhesive formed after the heating−cooling procedure. This might be ascribed to the rearrangement of PEO chain within the 3-D network structure of clay at high temperatures, leading to stronger cross-linking between PEO and clay which also accounts for reduced strain after cooling circle. Furthermore, during the whole procedure, the tackability of the dry gel also restored, as indicated by the recovery of tan δ value. 3.3. 3-D Network Structure of the NC Gels. For the dynamic restructuring and unusual adhesive properties, we propose that the reversible network structure formed by the interpenetrating of comb-like PEO chain within the amorphous 3-D silica and the weak physical interactions such as hydrogen bonding of the water with the polymer/clay, and possible

Figure 4. Frequency sweeps of the NC gels in fresh and dry states (A and B: red color, NC-gel-F1, 1 Pa; purple color, NC-gel-F2, 5 Pa; green color, NC-gel-F1P, 10 Pa; C: NC-F1T: blue color, 50 Pa; green color, 500 Pa; pink color, 1000 Pa).

NC-gel-F2 are dominant and frequency-independent, whereas loss moduli G″ increase with frequency. At higher frequencies, both moduli increase, with G″ rising more sharply and ultimately overtaking G′ at 1.1 and 10.8 Hz for NC-gel-F1 and NC-gel-F2, respectively. For the two dry gels, NC-gel-F1P and NC-F1T, both moduli increase nearly in parallel with frequency, with G′ remaining larger than G″ over the entire frequency range investigated. The frequency dependence also indicate that the fresh gels are viscoelastic solids (G′ > G″) at lower frequencies and viscoelastic liquids (G″ > G′) at higher frequencies, similar to globular actin filament network.18 In view of the study on a silica/PEO colloidal system in which E

dx.doi.org/10.1021/la401269p | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 7. XRD patterns of (a) pristine Na-MMT, (b) fresh PEO chemical hydrogel (for comparison purpose), (c) fresh NC-gel-F1, and (d) partially dried NC-gel-F1p.

Figure 5. Temperature sweeps of NC-FT at 1 Hz in a temperature range of 25−110 °C (blue, in a heating cycle; orange color, in a cooling cycle).

H-bonding of PEO with water, steric hindrance of 3D network structure of silica plates to the relaxation of polymer segments,24 and the coordination of PEO chains with sodium cations of clay might also account for the formation of amorphous PEO in the gel.25−28 The exfoliation of clay layers by polymers or/and silane was also confirmed by FTIR study. As shown in Figure 8A, the Si− O stretching vibrations were observed at 538 and 468 cm−1, showing the presence of the silica.29 Besides a strong band at 1038 cm−1 which corresponds to the Si−O vibrations of the tetrahedral sheet, relatively weaker absorptions at both 1088 and 800 cm−1 belong to that of amorphous silica with a threedimensional cross-linked silica framework.30 A comparison between Figures 8A1 and 8A2 indicates that the crystalline structure of pristine NaMMT was extensively transformed into amorphous structure by the chemical treatment of silane. 3.4. Hydrogen Bonding in the NC Gel. In Figure 8B, a broad peak in the range of 1690−1570 cm−1 can be deconvoluted into two components: one dominant at 1651 cm−1 corresponding to the bending vibration of water and the other at 1639 cm−1 attributed to that of water saturated lowdensity polyethylene or weakly bonded water molecules.31−33 The carbonyl stretching band at 1731 cm−1, which is sensitive to H-bonding, is relatively broader and asymmetric for PEG chemical hydrogel (Figure 8B1), suggesting possible hydrogen bonding interaction of ester groups with water. In NC hydrogels, this band becomes narrower and symmetric, indicating that this interaction might be interfered by the incorporation of clay (Figure 8B2). Peak values of IR spectra in the range of 800−1500 cm−1 for both pure amorphous PEG and PEG NC and chemical hydrogels are summarized in Table 1.31 By comparison, it can be concluded that only amorphous PEG exists in the both hydrogels. Of these absorption bands, the most significant difference between pure PEO and PEO in the NC gel is the C−O vibration stretching band which moves from 1140 to 1130 cm−1, possibly due to the formation of Hbonding between C−O−C of PEO chain with water or/and the clay.32 To confirm the existence of H-bonding among water and PEO/clay, ATR-FTIR was conducted in 4000−2500 cm−1 region which is more sensitive to H-bonding. Figure 8C shows that the stretching vibrational band of hydroxyl group of water increases in intensity and shifts to lower frequency in the NC gels with increasing water contents, indicating more water molecules participating in the formation of H-bonds among

polymer/cation complexation might be responsible. The possible structure of PEG NC hydrogel is illustrated in Figure 6.

Figure 6. Schematic illustration of the synthesis of PEG/clay nanocomposite (NC) hydrogel.

In Figure 7, in the lower 2θ range of 3°−10°, the complete disappearance of the silicate reflection in the XRD patterns of both fresh and partially dried PEG NC hydrogel, compared to pure NaMMT, suggesting the exfoliation of clay layers during the modification with TEPS and/or the synthesis of the NC hydrogel. Whereas in the 2θ range of 18°−35°, a broad peak appears for the NC and chemical hydrogels, instead of the characteristic sharp peaks at 19.1°, 23.2°, and 26.9° of pure crystalline PEG,21 implying that the structure of the polymer within the both hydrogels was amorphous. For the fresh NC and chemical gels, due to the high water contents in both gels, crystallization of PEO could be completely suppressed, leading to the exceptional ability of PEO to form hydrogen bonds with water in such an intricate and delicate way.22,23 Whereas in the partially dried NC gel NC-gel-F1P with much reduced water content, the absence of crystalline PEO suggests that other than F

dx.doi.org/10.1021/la401269p | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

water and PEO/clay. In contrast, this stretching band shifts to higher frequency with greatly decreased intensity for the fresh NC gel compared to that for the fresh chemical gel, with the both gels containing nearly the same water contents. This might be ascribed to topographic hindrance from 3-D network of clay. The types of interactions among the polymer, clay, and water within the NC and PEG chemical hydrogels were also examined by DSC, a sensitive technique to study thermotropic properties. Figure 9A shows the first heating curves of the both

Figure 9. DSC curves of NC-gel-F1 and NC-gel-FP and chemical PEG hydrogels (for comparison purpose): (A1) 1st heating of NC-gel-F1P, (A2) 1st heating of the dry chemical hydrogel, (A3) 1st heating of the NC-gel-F1, (A4) 1st heating of the fresh chemical hydrogel; (B1) 1st heating of the NC-gel-F1P, (B2) 1st cooling of the NC-gel-F1P, (B3) 2nd heating of the NC-gel-F1P, (B4) 1st cooling of the dry chemical gel.

Figure 8. FTIR spectra of (A1) pristine NaMMT, (A2) TEPSmodified NaMMT, (B1) fresh PEO chemical gel (for comparison purpose), (B2) partially dry NC hydrogel (NC-gel-F1P) and ATRFTIR spectra in 4000−2500 cm−1 region of (C1) fresh PEO chemical gel, (C2) fresh NC hydrogel (NC-gel-F1), and (C3) partially dry NC hydrogel (NC-gel-F1p).

gels in fresh and dry state from room temperature to 200 °C at a scan rate of 10 °C/min. For the fresh gels, a narrow endothermic peak appears at 97 and 115 °C for NC and chemical gels, respectively, whereas a broad endothermic peak of larger area at 136 °C for both. The narrow peak could be assigned to the water associated with the polymer and/or clay via H-bonding, whereas the broad peak probably belongs to the self-associated water. Two reasons are proposed for this assignment. One is based on the strength of hydrogen bonding which is inversely proportional to the distance of the H-bonds. In terms of this, self-associated water should have a H-bond of greater strength due to the shorter distance between small water molecules, whereas the steric hindrance around the polymer/clay might result in longer length of H-bond with water and thus a weaker strength of interassociated water. Therefore, compared to interassociated water, self-associated water being held by stronger hydrogen bonding forces require

Table 1. FTIR Characteristic Bands of Amorphous PEG and PEG NC and Chemical Hydrogels mode assignment EG CH2 scissor (gauche) EG CH2 wag (gauche) EG CH2 twist C−O, C−C stretch EG CH2 rocking (gauche) a

PEG amorphousa

NC and chemical hydrogels

1460 m

1456 m

1352 m 1296 m, 1249 m 1140 sh, 1107 s, 1038 m 945 m

1350 m 1291 m, 1244 m 1130 sh, 1106 s, 1033 m 947 m

Data cited from ref 31: m, medium; s, strong; sh, shoulder. G

dx.doi.org/10.1021/la401269p | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

4. CONCLUSIONS In summary, we developed NaMMT-based polymer nanocomposite hydrogels having both dynamic restructuring and unusual adhesive properties. They could be used as self-healing materials or find potential applications in acoustic dampers. The unique properties of our NC gels might be the result of a reversible 3-D network structure formed by comb-like PEO chains within amorphous silica. Our study could inspire more future research on the NaMMT/polymer nanocomposite hydrogels which might find potential applications in protein separation, drug and gene delivery, tissue engineering, and biomedical apparatus.

more thermal energy to remove, leading to a higher endothermic temperature.34 The releasing temperature of selfassociated water was also reported as high as above 150 °C due to the constraint of water in Laponite crystal structure as suggested by the producer of Laponite. Another reason is based on the total energy which is needed for the removal of different types of water. Since there is excess water (more than 80 wt %) contained in freshly prepared chemical and NC hydrogels, most of it should associate with each other, and the removal of it corresponds to larger area of the endothermic peak. At the same time, in Figure 6A, this peak remains at 136 °C for both the NC (fresh) and chemical gels (fresh and dry), indicating that the strength of self-associated water is hardly influenced within the different gels. Whereas for the interassociated water, the lower endothermic peak is located at different temperatures for the NC gel (fresh and dry) and chemical gel (fresh), implying the different H-bond strength of water interacting with either the polymer or the clay. Because of the relative weaking of interassociation, it is reasonable that the water should be removed easily first. This is the case in PEO chemical hydrogel as shown in Figure 9A2, in which the complete removal of interassociated water is indicated by the disappearance of the lower endothermic peak. In the NC hydrogel, it is the opposite. Self-associated water was totally removed upon drying at room temperature, whereas a small amount of interassociated water has been left, as suggested by the increase of the lower endothermic temperature to 109 °C from 97 °C with reduced peak intensity (Figure 9, A3 and A1). A possible explanation for this might be related to the different reversibility of network structure of the gels. Compared to the rigid chemical gel, the NC gel was linked by reversible and weaker physical interactions, and thus both the polymer and the cross-linker (clay) having higher mobility could form H-bonds with water repeatedly. Consequently, some water interacted with polymer/clay could remain after DSC heating, where selfassociated water might be removed completely during the first heating cycle. 3.5. Polymer/Cation Complexation in the NC Gel. Beside the first heating cycle, the first cooling cycle and the second heating cycle were also conducted on dry NC hydrogels at 10 °C/min and the DSC curves are shown in Figure 6B. On the first heating at 161 °C, there is another small and broad peak which reappears at 164 °C upon the second heating (Figure 9, B1and B3). This new endothermic temperature falls within a temperature range from 93 to 186 °C of the reversible dissolution of P(EO3Na+) complex observed in PEO−NaSCN and PEO−NaI systems.25−28 Thus, it seems that this PEO/Na+ complex also formed in our NC hydrogel. Upon the second heating, this peak decreases in intensity, whereas the endothermic peak of interassociated water shifts to 147 °C. Interestingly, there is an unexpected endothermic peak at 153 °C (Figure 6B2) upon cooling, which might be explained by some endothermic conversion occurring between PEO/water and PEO/Na+ complexes. However, for totally dry NC-gel-F1, the DSC curves during healing and cooling cycles are featureless, indicating that both the complexation of PEO and Na+ might not contribute to the restructuring and adhesive properties, whereas the 3-D network structure formed by PEO interpenetrating within amorphous silica is most likely to account for the reversible and resilient recovery of both the shape and strength of the dry gel.

■

ASSOCIATED CONTENT

S Supporting Information *

TGA curves for the determination of water contents of the NC hydrogel (NC-gel-F1 and NC-gel-F1P). This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (C.H.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Institute of Materials and Research Engineering, (IMRE), Agency for Science, Technology and Research (AStar), Singapore, for financial support.

■

ABBREVIATIONS PEG-OMe and PEG, poly(ethylene glycol) methyl ether acrylate; PEO, poly(ethylene glycol); NaMMT, sodium montmorillonite; NC, nanocomposite; ATR-FTIR, attenuated total reflectance Fourier transform infrared spectroscopy; DSC, differential scanning calorimetry; TGA, Thermal gravimetric analysis; XRD, X-ray diffraction.

■

REFERENCES

(1) Roberts, M. C.; Hanson, M. C.; Massey, A. P.; Karren, E. A.; Kiser, P. F. Dynamically restructuring hydrogel networks formed with reversible covalent crosslinks. Adv. Mater. 2007, 19, 2503−2507. (2) Olsen, B. D.; Kornfield, J. A.; Tirrell, D. A. Yielding behavior in injectable hydrogels from telechelic proteins. Macromolecules 2010, 43, 9094−9099. (3) Tziampazis, E.; Kohn, J.; Moghe, P. V. PEG-variant biomaterials as selectively adhesive protein templates: Model surfaces for controlled cell adhesion and migration. Biomaterials 2000, 21, 511−520. (4) Leach, J. B.; Schmidt, C. E. Characterization of protein release from photocrosslinkable hyaluronic acid-polyethylene glycol hydrogel tissue engineering scaffolds. Biomaterials 2005, 26, 125−135. (5) Hahn, M. S.; Taite, L. J.; Moon, J. J.; Rowland, M. C.; Ruffino, K. A.; West, J. L. Photolithographic patterning of polyethylene glycol hydrogels. Biomaterials 2006, 27, 2519−2524. (6) Almany, L.; Seliktar, D. Biosynthetic hydrogel scaffolds made from fibrinogen and polyethylene glycol for 3D cell cultures. Biomaterials 2005, 26, 2467−2477. (7) Yang, F.; Williams, C. G.; Wang, D.; Lee, H.; Manson, P. N.; Elisseeff, J. The effect of incorporating RGD adhesive peptide in polyethylene glycol diacrylate hydrogel on osteogenesis of bone marrow stromal cells. Biomaterials 2005, 26, 5991−5998. (8) Mahoney, M. J.; Anseth, K. S. Three-dimensional growth and function of neural tissue in degradable polyethylene glycol hydrogels. Biomaterials 2006, 27, 2265−2274.

H

dx.doi.org/10.1021/la401269p | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

(9) Chang, C. W.; Spreeuwel, A.; Zhang, C.; Varghese, S. PEG/clay nanocomposite hydrogel: a mechanically robust tissue engineering scaffold. Soft Matter 2010, 6, 5157−5164. (10) Wang, J.; Lin, L.; Cheng, Q.; Jiang, L. A strong bio-inspired layered PNIPAM−clay nanocomposite hydrogel. Angew. Chem. 2012, 124, 4754−4758. (11) Jiang, Y.; Li, B.; Chen, X.; Zhu, M. Preparation and characterization of a prolonged and sustained drug delivery system: Linear polyacrylamide in poly(N-isopropylacrylamide)/clay hydrogels. J. Appl. Polym. Sci. 2012, 125, E148−E156. (12) Wang, Q.; Mynar, J. L.; Yoshida, M.; Lee, E.; Lee, M.; Okuro, K.; Kinbara, K.; Aida, T. High-water-content mouldable hydrogels by mixing clay and a dendritic molecular binder. Nature 2010, 463, 339− 343. (13) Haraguchi, K.; Uyama, K.; Tanimoto, H. Self-healing in nanocomposite hydrogels. Macromol. Rapid Commun. 2011, 32, 1253−1258. (14) Krishnamoorti, R.; Ren, J.; Silva, A. S. Shear response of layered silicate nanocomposites. J. Chem. Phys. 2001, 114, 4968−4973. (15) Mezger, T. G. In Rheology Handbook: For Users of Rotational and Oscillatory Rheometers, 2nd ed.; Vincentz Network: Hannover, 2006; p 118. (16) Chu, S. G. In adhesive bonding: Dynamic Mechanical Properties of Pressure Sensitive Adhesives; Lee, H. L., Ed.; Plenum Pub. Corp.: New York, 1991; pp 31−45. (17) Demarteau, W.; Loutz, J. M. Rheology of acrylic dispersions for pressure sensitive adhesives. Prog. Org. Coat. 1996, 27, 33−44. (18) Xu, J.; Palmer, A.; Wirtz, D. Rheology and microrheology of semiflexible polymer solutions: actin filament networks. Macromolecules 1998, 31, 6486−6492. (19) Mason, T. G.; Weitz, D. A. Linear viscoelasticity of hard sphere suspensions near the glass transition. Phys. Rev. Lett. 1995, 75, 2770− 2773. (20) Krishnamoorti, R.; Giannelis, E. P. Rheology of end-tethered polymer layered silicate nanocomposites. Macromolecules 1997, 30, 4097−4102. (21) Wu, T.; He, Y.; Fan, Z.; Wei, J.; Li, S. Investigations on the morphology and melt crystallization of poly(L-lactide)-poly(ethylene glycol) diblock copolymers. Polym. Eng. Sci. 2008, 48, 425−433. (22) Philippova, O. E.; Kuchanov, S. I.; Topchieva, I. N.; Kabanov, V. A. Hydrogen bonds in dilute solutions of poly(ethylene glycol). Macromolecules 1985, 18, 1628−1633. (23) Zhang, J.; Han, F.; Wei, X.; Shui, L.; Gong, H.; Zhang, P. Spectral studies of hydrogen bonding and interaction in the absorption processes of sulfur dioxide in poly(ethylene glycol) 400 + water binary system. Ind. Eng. Chem. Res. 2010, 49, 2025−2030. (24) Jiang, S.; Yu, D.; Ji, X.; An, L.; Jiang, B. Confined crystallization behavior of PEO in silica networks. Polymer 2000, 41, 2041−2046. (25) Eastmond, G. C.; Schofield, P.; Sakellariou, P. Thermal properties of poly(ethylene oxide)-poly(methyl methacrylate) blends and copolymers complexed with sodium thiocyanate. Polymer 1997, 38, 1753−1761. (26) Robitaille, C.; Marques, S.; Boils, D.; Prud’homme, J. Thermal properties of poly(ethylene oxide) complexed with sodium thiocyanate and potassium thiocyanate. Macromolecules 1987, 20, 3023−3034. (27) Hague, J. D.; Wright, P. V. Semi-crystalline alkali-metal salt complexes with poly(oligooxyethyleneoxy-1,2-phenylene)s. J. Mater. Chem. 1991, 1, 507−509. (28) Strawhecker, K. E.; Manias, E. Crystallization behavior of poly(ethylene oxide) in the presence of Na+ montmorillonite fillers. Chem. Mater. 2003, 15, 844−849. (29) Madejova, J. FTIR techniques in clay mineral studies. Vib. Spectrosc. 2003, 31, 1−10. (30) Moenke, H. H. W. In Infrared Spectra of Minerals; Farmer, V. C. Ed.; Mineralogical Society: London, 1974; Chapter 16, p 365. (31) Harder, P.; Grunze, M.; Dahint, R. Molecular conformation in oligo(ethylene glycol)-terminated self-assembled monolayers on gold and silver surfaces determines their ability to resist protein adsorption. J. Phys. Chem. B 1998, 102, 426−436.

(32) Lele, A. K.; Badiger, M. V. Thermodynamic of hydrogen-bonded polymer gel-solvent systems. Chem. Eng. Sci. 1995, 50, 3535−3545. (33) Manoratne, C. H.; Rajapakse, R. M. G.; Dissanayake, M. A. K. L. Ionic conductivity of poly(ethylene oxide) (PEO)-montmorillonite (MMT) nanocomposites prepared by intercalation from aqueous medium. Int. J. Electrochem. Sci. 2006, 1, 32−46. (34) Bhunia, M. K.; Das, S. K.; Bhaumik, A. Temperature induced proton transfer in a hydrogen bonded supramolecule. Chem. Phys. Lett. 2010, 498, 145−150.

I

dx.doi.org/10.1021/la401269p | Langmuir XXXX, XXX, XXX−XXX