3D Cocontinuous Composites of Hydrophilic and Hydrophobic Soft


3D Cocontinuous Composites of Hydrophilic and Hydrophobic Soft...

2 downloads 79 Views 4MB Size

Letter pubs.acs.org/macroletters

3D Cocontinuous Composites of Hydrophilic and Hydrophobic Soft Materials: High Modulus and Fast Actuation Time Junseok Kim, Yunho Cho, Soyun Kim, and Jonghwi Lee* Department of Chemical Engineering and Materials Science, Chung-Ang University, 221, Heukseok-dong, Dongjak-gu, Seoul, 156-756, Republic of Korea S Supporting Information *

ABSTRACT: Hydrogels in nature seldom form a single phase, more often forming structured phases with other soft phases, allowing nature to develop responsive and adaptive strategies. Based on knowledge of how hydrogels are utilized in nature, we developed novel 3D cocontinuous composites from soft materials with extremely different properties, a hydrogel and a silicone. These were successfully prepared by infiltrating liquid polydimethylsiloxane (PDMS) into poly(N-isopropylacrylamide) (PNIPAm) frameworks of aligned pores prepared by directional melt crystallization. The composites had outstanding modulus and swelling ratio compared to other mechanically strong hydrogels. More interestingly, the deswelling kinetics were dramatically accelerated (by a factor of 1000), possibly due to the aligned microchannels and the hydrophobic nature of PDMS. As a result, an actuator movement mimicking flowering could be completed in less than 20 s. This novel and versatile cocontinuous composite strategy could overcome the current limitations of soft materials.

N

bottleneck because only a few techniques such as timeconsuming 3D printing are currently available.9,12 In this study, we introduce an efficient and versatile novel composite preparation technique based on the directional melt crystallization of a solvent (also known as ice templating, directional freezing, and freeze-casting),13−16 resulting in pores with 3D connectivity and low tortuosity after removing the crystals.17 The pore morphology replicated the crystal habit of ice, as the detailed mechanisms can be found elsewhere.13−18 When NIPAm solution was brought in contact with liquid nitrogen, ice crystals nucleated from the bottom of the solution and grew to the top surface (Figure 1a and SI). The relatively wellcontrolled unidirectional temperature gradient guided the growth of the ice crystals, and subsequent cross-linking in cryoconcentrate phases fixed the structures before removal of the crystals.18 Our experimental conditions, such as freezing rate, produced a pore morphology with cylindrical or lamellar characteristics in a fishbone structure in which the pores were highly connected, which is consistent with previous publications.17,18 Isolated homogeneous nucleation, if occurs, will produce closed pores. However, nature prefers heterogeneous nucleation more, which makes this crystallization-based method produce unique continuous open pores.17 The conventional method of pore preparation using foaming agents or soluble particles also usually creates closed pores.19,20 The 3D

atural soft materials have been found to have tremendously high variation in physicochemical properties and microstructures.1,2 On the other hand, artificial soft materials suffer from limited properties and seldom employ the sophisticated architectures found in nature.3 Numerous soft materials have been developed over recent decades, but structural combinations of them have been limited.4 For example, hydrogels, a typical hydrophilic soft material, and silicones, a typical hydrophobic soft material, have seldom been combined into composites with functional architectures of separate and distinct phases.5−8 Their extremely different properties have made the typical composite preparation of infiltration or blending difficult, while the preparation of interpenetrating network, hybrid or conjugate materials are straightforward. Nevertheless, the development of a method for producing composites of the two extremes could bring us an entirely novel class of soft materials with unique properties. To prove this, we herein developed a general method using cocontinuous structures. In a cocontinuous composite, each phase forms a continuous network that is penetrated by similar networks of other phases.9−11 This is an attractive solution, because it provides freedom from interfacial strength concern, while allowing the composite to retain the properties of each component phase. Despite of poor interfacial strength between silicones and hydrogels, cocontinuous structure allows us to combine them with retaining the good mechanical properties of silicones and the water compatibility of hydrogels. Cocontinuous composites have advanced in parallel with the development of their preparation technique, which is the main © XXXX American Chemical Society

Received: August 22, 2017 Accepted: September 26, 2017

1119

DOI: 10.1021/acsmacrolett.7b00642 ACS Macro Lett. 2017, 6, 1119−1123

Letter

ACS Macro Letters

shrinking (Figure S1). The crystallization rate, C, is directly related to freezing rate, and L is known to be expressed by the following relationship, based on material balance and Fick’s law: L = (8DΔT )0.5 C −0.5

(1)

where ΔT is the supercooling between lamellae, and D is the binary diffusion coefficient.17 Although this equation is too simple to quantitatively explain the relationship between L and C, the relationship is qualitatively consistent with our results. Phase boundaries became distinct after swelling (Figure S2), and the unique cylindrical or lamellar structures of composites were identified. Freeze-drying after swelling in 4 °C water revealed the expanded porous phases of PNIPAm, which could be clearly identified on the cross sections and surfaces of the composites (Figure 1f,g). The protruded PNIPAm phases are evidence of volume expansion due to swelling. As shown in Figure 1f, the average width of the PDMS phase is 47.2 ± 18.4 μm, which is similar to the value of L of Comp-f (pore spacing 50.7 μm). Therefore, it can be stated that PDMS was successfully infiltrated into the pores of PNIPAm. The average width of the PNIPAm was 39.7 ± 16.4 μm. The identification of phases was also confirmed by energy dispersive X-ray spectroscopy (EDX). Cocontinuous composites of hydrogels and silicones are able to satisfactorily combine the mechanical properties of commercial rubbers with the water-swelling properties of hydrogels. The stress−strain curves of the two materials are extremely different, showing the characteristics typical of commercial hydrophobic rubbers and hydrogels in water (difference in modulus of more than 2 orders of magnitude). The moduli of the composites were between those of PDMS and PNIPAm (Figure 2a,b and Table S1). The compressive and tensile moduli of the composites (0.12−0.63 MPa) are comparable to those of commercial soft rubbers. Their relatively high moduli along with their water retention properties are attractive advantages of these cocontinuous composites.21 The pore alignment induced anisotropy in the mechanical properties. In general, columnar or lamellar pores result in higher moduli along the pore direction. It is also known that macromolecular chains align along the pore direction during directional melt crystallization or infiltration.16 As a result, the moduli in the direction parallel to the pores are higher than those in the perpendicular direction. The mechanical properties of hydrogels are predominantly dependent on the swelling ratio. As the swelling ratio of a hydrogel decreases, the elastic modulus increases. The scaling law derived from the simple rubber elasticity theory relates the elastic modulus of the hydrogel, E, to the volume fraction of the polymer, φ, as follows:22

Figure 1. (a) Porous PNIPAm hydrogel prepared by the directional melt crystallization technique. (b) Three structures of PNIPAmPDMS composites obtained by infiltration. (c) OM and fluorescence OM, (d) Micro-CT, and (e) SEM images of porous PNIPAm. (f) Cross-sectional SEM and EDX images (Si mapping) of Comp-f, freeze-dried after swelling. (g) Scanning electron microscope (SEM) image of the top surface of Comp-f. (⊥) Surfaces perpendicular to pore direction; (∥) surfaces parallel to pore direction; (↑) direction of crystal growth.

continuous and aligned pore structure allowed for easy infiltration of uncured liquid PDMS, and the cross-linked PNIPAm structure provided dimensional stability during the infiltration (Figure 1b). The porous PNIPAm plays the role of “preform” for the infiltration step of composites, and the resulting composites had unique 3D cocontinuous structures of separate microphases. One composite film (Comp-f) and two composite cylinders (Comp-c and Comp-hc) were prepared (SI). The 3D continuous and aligned pores running through the thickness of the structure were revealed using optical microscopy (OM) and fluorescence microscopy (Figure 1c). Micro computed tomography (micro-CT) further confirmed the 3D continuity of the pores, which was critical for the infiltration of PDMS liquid (Figure 1d and Movie 1). The pores were well aligned with the temperature gradient (pore direction) and formed a regular pattern of fishbone-structured holes perpendicular to the temperature gradient (Figure 1e). The spacing between pore walls, L, was measured to be 50.7 ± 16.2 μm for Comp-f (lower freezing rate) and 17.7 ± 5.2 μm for Comp-c (Figures 1e and S1). The use of a porous PNIPAm cylinder shrunk in 40 °C water resulted in a contracted and wrinkled pore structure, with L = 11.3 ± 5.7 μm (Comp-hc). The 3D continuous pore structure was maintained after

E ≅ Tφ 2.25(v 3/4a3/2)

(2)

where T is temperature, v is the excluded volume parameter, and a is the effective length per monomer. The swelling ratio is inversely proportional to the elastic modulus. Therefore, for a more rational comparison of moduli, the swelling ratio should also be considered. In Figure 2c,d, our plots of modulus versus swelling ratio are compared with those of previously reported strong hydrogels. These previous studies used various strengthening methods. Compared to the existing strengthening methods, our cocontinuous composite method produced 1120

DOI: 10.1021/acsmacrolett.7b00642 ACS Macro Lett. 2017, 6, 1119−1123

Letter

ACS Macro Letters

Figure 2. Stress−strain behavior in (a) uniaxial compressive tests and (b) uniaxial tensile tests (⊥) perpendicular to the pore direction and (∥) parallel to the pore direction. (c) Compressive modulus vs swelling ratio. (d) Tensile modulus vs swelling ratio. (IP: hydrogels with inorganic particles; OP: hydrogels with organic particles; DN: double network hydrogels; IPN: interpenetrating polymer network hydrogels; references in Tables S2 and S3).

Figure 3. (a) Normalized equilibrium swelling ratio vs temperature. (b) Swelling kinetics in 4 °C water. (c) Deswelling kinetics in 4 °C water. (d) Effective diffusion coefficient vs tensile modulus is plotted to compare the intrinsic actuation performances (NN: nanostructured network hydrogels. References in Table S4).

1121

DOI: 10.1021/acsmacrolett.7b00642 ACS Macro Lett. 2017, 6, 1119−1123

Letter

ACS Macro Letters

with low tortuosity (aligned microchannels) facilitating water diffusion. Recently, it was reported that the deswelling rate of a PNIPAm/PDMS copolymer was increased by the hexagonal (channel) structure of PNIPAm obtained using liquid crystals.25 Furthermore, the hydrophobic surfaces of PDMS can also increase the deswelling kinetics of the composites. The superfast response is particularly surprising because the composites have smaller open surface area for water diffusion due to the presence of PDMS surface layers. The superfast response of the composites allows them to follow cyclic temperature stimuli and cyclic release control (Figure S4 and Movie 2). Comp-f, which has only one exposed PNIPAm surface, can produce a bending motion by swelling and deswelling. This repeatable fast response can be utilized for reversible flowering and withering movements (Figure 4a and

competitive and even superior results, particularly in terms of tensile modulus versus swelling ratio. The volume transition behavior, which is based on the lower critical solution temperature (LCST) of PNIPAm, was investigated by measuring the equilibrium swelling ratio (ESR) of the composites (Figure 3a and SI). Because of the spatial arrangement of PNIPAm and PDMS, the composites achieved up to 200% ESR in 4 °C water, even though they contained 80 wt % PDMS. Even at 200% ESR, the moduli of the composites were surprisingly high compared with those of conventional hydrogels; the composites had tensile and compressive moduli that were 50× and 36× higher than those of PNIPAm, respectively. Both PNIPAm and the composites exhibited a distinct decrease in swelling ratio around 32 °C, and the composites had the same temperature responsive characteristics. As shown in Figure 3b, when Comp-f was immersed in 4 °C water, it swelled as fast as PINPAm. Both films had the same thickness (1 mm). On the other hand, Comp-c was cylindrical, with a thickness and diameter (10 mm) much larger than the thickness of the films. This resulted in slower volume expansion upon swelling. Comp-hc, which was prepared from shrunken PNIPAm, showed much slower swelling kinetics, possibly due to the dimensional difference as well as the smaller expansion strain produced by shrunken PNIPAm. The shrunken pores of Comp-hc have a smaller diameter and are more serpentine water diffusion channels, delaying the swelling kinetics. A remarkable and unique property of the composites was noticed in the deswelling experiment. The time for PNIPAm to shrink to 10% of its full swelling volume was 5 min (inset of Figure 3c). Surprisingly, the composites deswelled much faster than PNIPAm. Comp-f with the same dimensions as PNIPAm deswelled to 10% of its initial volume in only 12 s (Figure 3c). Comp-c and Comp-hc required only 44 and 63 s, respectively, even with a thickness of 10 mm, 10 times thicker than that of PNIPAm. Therefore, the water content and volume of the composites changed with temperature on the order of seconds, not minutes. The volume transition should be accompanied by water diffusion into or out of the composites (weight changes); this was also experimentally confirmed (Figure S3). The volume transition results are converted to effective diffusion coefficients.23 The effective diffusion coefficients of Comp-f (735 × 10−6 cm2/s) and Comp-c (781 × 10−6 cm2/s) are 1000× larger than that of PNIPAm (0.765 × 10−6 cm2/s). The similarity between the coefficients of Comp-f and Comp-c indicates that they have similar internal structures and similar water diffusion microprocesses, even though they have different curves in Figure 3 due to different specimen dimensions. The effective diffusion coefficient represents the capacity for responsive actuator motion. However, to produce useful mechanical motion, mechanical properties such as the modulus should also be considered. Figure 3d shows a plot of the effective diffusion coefficient vs the tensile modulus of our composites and other hydrogels.24 Compared to the other hydrogels, whose properties were reported in previous literature, our composites have not only higher effective diffusion coefficients, but also superior tensile moduli. The superfast responsive behavior can be explained by the 3D cocontinuous structure of the composites and the hydrophobic nature of PDMS. The aligned porous structure of PNIPAm is incorporated into the composites, providing 3D continuity and alignment of separate and distinct PNIPAm phases. These phases act as organized water diffusion channels

Figure 4. (a) Comp-f showing fast volume transition upon contact with 4 °C (upper) and 40 °C (lower) water. SEM cross-sectional images of Comp-f swollen below and above LCST: (b) 4 °C and (c) 40 °C.

Movie 3). Both flowering and withering take less than 20 s, and withering in particular takes less than 10 s. These characteristics show the potential for use as relatively fast responsive actuators for robot hands or drug release control.26,27 The structure of the composites in water above and below LCST was analyzed using cross-sectional surfaces (Figure 4b,c). Below LCST, there was internal stress originating from the expansion of the PNIPAm phase, which was confirmed by its protruded structures. On the other hand, shrunken PNIPAm phases produced opposite stress above LCST. As a result, the average width of the PNIPAm phase below and above LCST was 39.7 ± 16.4 and 12.1 ± 6.5 μm, respectively. PDMS compresses when PNIPAm swells and elastically recovers its original volume when PNIPAm shrinks. Through this process, PDMS could compress the PNIPAm phases and accelerate the diffusion of water from the composites. In summary, we successfully fabricated composites of PNIPAm and PDMS by infiltrating PDMS into the continuous pores of PNIPAm. Hydrophobic and hydrophilic phases with extremely different properties coexisted in the 3D cocontinuous composites. The composites absorbed water like a hydrogel and had improved mechanical properties like a silicone. These properties were outstanding compared to previously reported strong hydrogels. Due to their 3D cocontinuously aligned structure, the deswelling rate of the composites dramatically increased. Their highly temperature-sensitive properties 1122

DOI: 10.1021/acsmacrolett.7b00642 ACS Macro Lett. 2017, 6, 1119−1123

Letter

ACS Macro Letters

(15) Lee, M. K.; Rich, M. H.; Shkumatov, A.; Jeong, J. H.; Boppart, M. D.; Bashir, R.; Gillette, M. U.; Lee, J.; Kong, H. Adv. Healthcare Mater. 2015, 4 (2), 195−201. (16) Lee, M. K.; Lee, J. Nanoscale 2014, 6 (15), 8642−8648. (17) Kim, B. S.; Lee, J. Chem. Eng. J. 2016, 301, 158−165. (18) Halake, K. S.; Lee, J. Carbohydr. Polym. 2014, 105, 184−192. (19) Choi, S.-J.; Kwon, T.-H.; Im, H.; Moon, D.-I.; Baek, D. J.; Seol, M.-L.; Duarte, J. P.; Choi, Y.-K. ACS Appl. Mater. Interfaces 2011, 3 (12), 4552−4556. (20) Haugen, H.; Ried, V.; Brunner, M.; Will, J.; Wintermantel, E. J. Mater. Sci.: Mater. Med. 2004, 15 (4), 343−346. (21) Ma, W.-S.; Li, J.; Zhao, X.-S. J. Mater. Sci. 2013, 48 (15), 5287− 5294. (22) Gennes, P.-G. d. Scaling Concepts in Polymer Physics; Cornell University Press, 1979. (23) Gao, P.; Nixon, P. R.; Skoug, J. W. Pharm. Res. 1995, 12 (7), 965−971. (24) Xia, L.-W.; Xie, R.; Ju, X.-J.; Wang, W.; Chen, Q.; Chu, L.-Y. Nat. Commun. 2013, 4, na. (25) Forney, B. S.; Baguenard, C.; Guymon, C. A. Soft Matter 2013, 9 (31), 7458−7467. (26) Liu, Y.; Zhang, K.; Ma, J.; Vancso, G. J. ACS Appl. Mater. Interfaces 2017, 9, 901−908. (27) Li, X.; Cai, X.; Gao, Y.; Serpe, M. J. J. Mater. Chem. B 2017, 5, 2804−2812.

enabled them to respond in seconds. These beneficial properties, that is, high modulus, high swelling ratio, and fast volume response, demonstrate the potential of these composites in broader applications such as pulsatile drug release, touch sensors, actuators, lap-on-a-chip, cooling systems, wound healing, antifouling, and wettable silicones.



ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsmacrolett.7b00642. Experimental details and additional figures and tables (PDF). Movie S1 (AVI). Movie S2 (AVI). Movie S3 (AVI). Movie S4 (AVI). Movie S5 (AVI).



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Jonghwi Lee: 0000-0003-2336-8695 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by a National Research Foundation Grant from the Korea Ministry of Science, ICT, and Future Planning (Engineering Research Center 2014R1A5A1009799 and 2016R1A2B4011247).



REFERENCES

(1) Wisedchaisri, G.; Reichow, S. L.; Gonen, T. Structure 2011, 19 (10), 1381−1393. (2) Lee, J.; Macosko, C. W.; Urry, D. W. Macromolecules 2001, 34 (17), 5968−5974. (3) Annabi, N.; Tamayol, A.; Uquillas, J. A.; Akbari, M.; Bertassoni, L. E.; Cha, C.; Camci-Unal, G.; Dokmeci, M. R.; Peppas, N. A.; Khademhosseini, A. Adv. Mater. 2014, 26 (1), 85−124. (4) Gong, J. P. Science 2014, 344 (6180), 161−162. (5) Cha, C.; Antoniadou, E.; Lee, M.; Jeong, J. H.; Ahmed, W. W.; Saif, T. A.; Boppart, S. A.; Kong, H. Angew. Chem., Int. Ed. 2013, 52 (27), 6949−6952. (6) Cui, J.; Lackey, M. A.; Madkour, A. E.; Saffer, E. M.; Griffin, D. M.; Bhatia, S. R.; Crosby, A. J.; Tew, G. N. Biomacromolecules 2012, 13, 584−588. (7) Cui, J.; Lackey, M. A.; Tew, G. N.; Crosby, A. J. Macromolecules 2012, 45, 6104−6110. (8) Gabor, E.; Kennedy, J. P. Prog. Polym. Sci. 2006, 31, 1−18. (9) Wang, L.; Lau, J.; Thomas, E. L.; Boyce, M. C. Adv. Mater. 2011, 23 (13), 1524−1529. (10) Breslin, M.; Ringnalda, J.; Xu, L.; Fuller, M.; Seeger, J.; Daehn, G.; Otani, T.; Fraser, H. Mater. Sci. Eng., A 1995, 195, 113−119. (11) Chen, J.; Cui, X.; Zhu, Y.; Jiang, W.; Sui, K. Carbon 2017, 114, 441−448. (12) Liu, Q.; Ye, F.; Gao, Y.; Liu, S.; Yang, H.; Zhou, Z. J. Alloys Compd. 2014, 585, 146−153. (13) Zhang, H.; Hussain, I.; Brust, M.; Butler, M. F.; Rannard, S. P.; Cooper, A. I. Nat. Mater. 2005, 4 (10), 787−793. (14) Gao, H. L.; Xu, L.; Long, F.; Pan, Z.; Du, Y. X.; Lu, Y.; Ge, J.; Yu, S. H. Angew. Chem., Int. Ed. 2014, 53 (18), 4561−4566. 1123

DOI: 10.1021/acsmacrolett.7b00642 ACS Macro Lett. 2017, 6, 1119−1123