Gyrification-Inspired Highly Convoluted Graphene Oxide Patterns for Ultralarge Deforming Actuators Yinlong Tan, Zengyong Chu,* Zhenhua Jiang, Tianjiao Hu, Gongyi Li, and Jia Song College of Science, National University of Defense Technology, Changsha 410073, P. R. China S Supporting Information *
ABSTRACT: Gyrification in the human brain is driven by the compressive stress induced by the tangential expansion of the cortical layer, while similar topographies can also be induced by the tangential shrinkage of the spherical substrate. Herein we introduce a simple three-dimensional (3D) shrinking method to generate the cortex-like patterns using two-dimensional (2D) graphene oxide (GO) as the building blocks. By rotation-dip-coating a GO film on an air-charged latex balloon and then releasing the air slowly, a highly folded hydrophobic GO surface can be induced. Wrinkling-to-folding transition was observed and the folding state can be easily regulated by varying the prestrain of the substrate and the thickness of the GO film. Driven by the residue stresses stored in the system, sheet-to-tube actuating occurs rapidly once the bilayer system is cut into slices. In response to some organic solvents, however, the square bilayer actuator exhibits excellent reversible, bidirectional, largedeformational curling properties on wetting and drying. An ultralarge curvature of 2.75 mm−1 was observed within 18 s from the original negative bending to the final positive bending in response to tetrahydrofuran (THF). In addition to a mechanical hand, a swimming worm, a smart package, a bionic mimosa, and two bionic flowers, a crude oil collector has been designed and demonstrated, aided by the superhydrophobic and superoleophilic modified GO surface and the solvent-responsive bilayer system. KEYWORDS: bilayer actuator, graphene oxide, gyrification, 3D shrinking, solvent responsive, bionic mimosa, crude oil collector
T
after multistep manipulation. Therefore, it is crucial to develop a facile one-step and transfer-free approach to building highly buckling patterns with hierarchical microstructures on spherical curved substrates, which may widen their applications, for example, in the fields of photodetectors,18,19 microbarriers, and nanoreactors.20,21 In nature, however, there are indeed diverse buckling patterns on spherical substrates, such as wrinkling patterns on cells, folding patterns of human cerebral cortex, and other patterns on spherical or oval fruits.15,22,23 A number of studies have demonstrated that the formation of these patterns in natural systems is partly due to mechanical instability between the comparatively rigid shell and soft core substrate.23,24 For example, gyrification in the human brain is an efficient way of packing a large surface area cerebral cortex into a relatively small volume provided by the human skull.23 The peak of such a fold is called a gyrus, and its trough is called a sulcus. Even though the physiological
he fabrication of micro/nano-patterns is of great significance to material science and technology,1,2 with numerous potential applications in microfluidics and microimprinting,3,4 wetting and adhesion,5,6 surface-enhanced Raman scattering (SERS),7 flexible electronics,8,9 mechanical property measurements,10,11 and cell culture biointerfaces.12 As a low-cost and highly efficient method, mechanical-driven buckling patterns including wrinkles, creases, and folds of thin films on compliant substrates have been intensively investigated.13,14 Although considerable progress has been made in both experimental fabrication and theoretical analysis of the buckling patterns, there remain some challenges: (i) Most previous studies mainly focus on the mechanical self-assembly on planar substrates while the experimental fabrication of threedimensional (3D) micro/nano-patterns on curved substrates was rarely reported.15 Among the limited studies on the curved substrates, the low mismatched strain between the film and the substrate limits the variation of the patterns.15,16 (ii) The fabrication of multiscale topographies remains complicated since sequential mechanical deformation or sequential transfer of the films are needed during the fabrication process,2,17 which makes it difficult to control the deformation or maintain the patterns © 2017 American Chemical Society
Received: March 21, 2017 Accepted: June 5, 2017 Published: June 5, 2017 6843
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
Article
www.acsnano.org
Article
ACS Nano
Scheme 1. Schematic Illustration of the Fabrication Process To Generate Highly Folded Cortex-like Patterns on a Latex Balloon Substratea
a
(a) 3D illustration of the samples obtained at different stages. A model of the human brain is inserted at the last step which is the key stage to mimic the gyrification process. (b) Top view illustration indicating the volume-shrinking-induced compressive stresses and highly folded surface topography. (c) Cross-sectional view illustration indicating the tangential compressive stresses and cross-sectional topography. For clarity, sizes are represented not in proportion. The diameters of the balloons (d1, d2, d3, d4), the thickness of the GO film (hf), the tensile pre-strain (εpre‑s), and the residue tensile strain (εres‑s) of the substrate, the compressive strain (εf) of the GO film, the typical width (W), and the maximum depth (Dm) of the folded GO ridges are labeled at necessary positions.
The deformation comes from the competition of the residue compressive stress stored on the highly folded GO side and the swelling-induced tensile stress on the latex side. An ultralarge and reversible curvature of 2.75 mm−1 was observed in the study.
mechanism behind gyrification has been unclear, at present, the most likely hypothesis is the simplest: tangential expansion of the cortical layer generates compressive stress, leading to the mechanical folding of the cortex.24,25 Very recently, Tallinen and co-workers successfully mimicked the gyrification process using a 3D-printed layered gel with real fetal brain geometry,26 where the compressive stress is initiated by the swelling-expansion of the surface layer. Since compressive stress can be either initiated by the volume expansion of the surface layer, or by the volume shrinkage of the substrate,27 herein we report a more general, highly efficient one-step 3D shrinking method to generate highly folded patterns capable of dynamic tunability. We select air-charged latex balloon as the 3D soft substrate, and graphene oxide (GO) as the relatively rigid surface layer. GO is a twodimensional (2D) soft but strong atomic sheet with oxygen containing groups, commercially available in large quantities and highly dispersible in water.14 The GO layer can be easily modified or reduced in situ to adjust the hydrophobic and oleophilic properties.17 Simultaneous 3D-shrinking is easily realized by releasing the air slowly and the compressive stress is meanwhile induced on the GO layer. Beyond a critical point, highly folded patterns analog the cerebral cortex can be generated, whose topographies can be tuned by controlling the diameter of the air-charged balloon (i.e., the tensile prestrain of the substrate) as well as the thickness of the film. Actuators are the materials and devices that are able to change their shapes in response to various changes in the environmental conditions and thus perform mechanical work on the nano-, micro-, and macro-scales.28,29 In recent years, various kinds of gaphene-based asymmetric bilayer materials have been demonstrated as manipulators, walkers, and drug carriers.30−32 However, stimuli-responsive bilayer actuators with reversible large curvatures, fast responding rates, and programmable bidirectional deformations are still critical for their practical applications.33 Here we also demonstrate the actuation property of the highly folded-GO/latex bilayer composite. Interestingly, the bilayer actuator exhibits reversible highly bending property in response to some organic solvents, such as n-hexane, benzene, and tetrahydrofuran (THF).
RESULTS AND DISCUSSION Scheme 1 illustrates the rotation-dip-coating and 3D-shrinking process to produce highly folded GO patterns using a commercially available spherical latex balloon as the substrate. More details about the fabrication process are shown in Supporting Information Figure S1. An aqueous GO suspension is rotationdip-coated and air-dried on the surface of an air-charged balloon, forming a thin GO coating. The thickness of the GO coating can be regulated from 0.5 to 2.5 μm by varying the casting concentration of the GO suspension from 2.0 mg/mL to 7.0 mg/mL (Supporting Information Figure S2). There are plenty of rivet-like protrusions on the rough surface of the balloon substrate, which provides advantageous strong interactions with GO (Supporting Information Figure S3). GO could smooth the balloon surface forming a uniform film with first-generation wrinkles. Subsequent to the coating process, the tangential tensile prestrain in the balloon was released three-dimensionally by discharging the air slowly, leading to a volume-shrinking-induced compressive stress on the GO film (Scheme 1b,c). As a result, the film was forced to buckle into highly convoluted cortex-like patterns so as to pack its large surface into a much limited area. For simplicity, the diameters of the initial balloon, the air-charged balloon, the air-charged GO/latex bilayer system, and the air-discharged bilayer system are denoted as d1, d2, d3, and d4, respectively, in which d3 = d2 + 2hf ≈ d2 since d2≫ hf. Here we define the tangential tensile prestrain of the substrate and the tangential compressive strain of the GO film as εpre‑s = (d2-d1)/d1 and εf = (d3-d4)/d3 ≈ (d2-d4)/d2, respectively. In general, a residue tensile strain exists in the air-discharged bilayer system due to a higher value of d4 than that of d1, which is defined as εres‑s = (d4-d1)/d1. Since the degree of folding is conventionally quantified by the gyrification index (GI), the 6844
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
Article
ACS Nano
Figure 1. Surface and cross-sectional SEM images of the buckled GO films with variation of the tensile prestrains of the substrate. (a) εpre‑s = 22%, (b) εpre‑s = 100%, (c) εpre‑s = 262%, and (d) εpre‑s = 378%. Compressive strains of the GO film are also listed as εf = 11, 38, 59, and 67%, respectively, whose thickness is controlled as ∼2.0 μm. The substrates are dyed with the color of mulberry. The red arrows on the surfaces indicate the meeting point of several ridges. The values of the typical width (W), the maximum depth (Dm) of the folded GO ridges, and the thickness of the final substrate (hs) are labeled at necessary positions.
mismatched strains between the film and the substrate in the previous studies reduced their opportunities to generate folding patterns.15,34 But in this study, hierarchical folding patterns on spherical substrates can be successfully obtained by increasing the prestrains remarkably up close to 400% (Figure 1d). Magnified views of the surface topographies in Figure 1b−d reveal detailed cortex-like microstructures on the GO film. The firstgeneration wrinkles can also be observed on the surface of the ridges. Cross-sectional images of the bilayer system in Figure 1 illustrate that the thickness of the balloon substrate decreases with the increase of the tensile prestrain, confirming that there are increased residual tensile strains in the balloon substrate. The residual tensile strain is the largest, εres‑s = 60%, for the sample prepared with the largest tensile prestrain (εpre‑s = 378%) and the largest gyrification index (GI = 8.9) (Supporting Information, Table S1). In addition, the width of the ridges (W) reduces remarkably with increasing of the prestrains, and the maximum depth (Dm) increases simultaneously (Supporting Information, Figure S5). It means that the interaction between the GO film and the substrate increases with increasing of the prestrains, which is associated the formation of strongly convoluted numerous microscale ridges (“gyri”) and trenches (“sulci”).
ratio of the surface area to the area of the convex hull,25,26 here GI is calculated as GI = (d3/ d4)2 ≈ (d2/ d4)2. By varying the air-charged balloon, d2, the tangential tensile prestrain of the substrate can be tuned from 22% to 378%, meanwhile the tangential compressive strain of the GO film can be regulated from 11% to 67% (Supporting Information, Table S1). Accordingly, GI is changing from 1.3 to 8.9. Such a high gyrification index makes it possible to generate highly folded cortex-like patterns on the GO film. In addition, we assign W and Dm as the typical width and the maximum depth of the folded GO ridges, so as to characterize the folding state with these feature sizes, which have been labeled in Scheme 1b,c. First we investigated the buckling patterns and the morphological evolution in the GO film with the variation of the tensile prestrains at a controlled thickness. As shown in Figure 1, SEM images in Figure 1a indicate that the GO film tends to shrink into a buckyball-like pattern at a comparatively lower tensile strain (εpre‑s = 22%, GI = 1.3). Similar topography can also be observed on a tangerine surface during the dehydration process (Supporting Information, Figure S4). With the increase of the tensile prestrain up to 100% and beyond, a wrinkle-to-ridge transition occurs, leading to the formation of cortex-like microstructures (Figure 1b−d). It is worth mentioning that small 6845
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
Article
ACS Nano
Figure 2. Surface and cross-sectional SEM images of the folded GO films with variation of the thickness of the GO film. (a) hf = 0.5 μm, (b) hf = 0.7 μm, (c) hf = 1.4 μm, and (d) hf = 2.5 μm. The tensile prestrain of the substrate is controlled as ∼140%. The substrates are dyed with the color of mulberry. The values of the typical width (W), the maximum depth (Dm) of the folded GO ridges, and the thickness of the final substrate (hs) are labeled at necessary positions.
Effect of the thickness of GO film on the morphology evolution was further studied by controlling the tensile prestrain as ∼140%, as shown in Figure 2. At this prestrain, all the samples exhibit highly folded topographies, but both the width of the ridges and the maximum depth of the folded GO film experience an upward trend with increasing the thickness of the films from 0.5 to 2.5 μm (Supporting Information, Figure S5).
The mechanical mechanism behind the 3D buckling has been recently analyzed from the stress or energy viewpoints.27,35 As the tangential prestrains in the spherical substrate are released, the film is forced to shrink under the tangential compressive stress, σf. When the compressive stress σf in the film exceeds a critical value, the formation of buckling patterns can be induced. The compressive stress σf can be calculated as35
where Es, R, and νs are the shear modulus, radius, and Poisson’s ratio of the spherical substrate while Ef, hf, and νf represent the shear modulus, thickness, and Poisson’s ratio of the film. Δε is the mismatched strain between the substrate and the film. Equation 1 reveals that some key parameters such as Δε and hf can be used to tune the topographies of buckling patterns. As predicted by Li and co-workers,27 with increasing deformation, the sphere first exhibits a buckyball-like wrinkling pattern and then undergoes a wrinkle-to-fold transition leading to a labyrinth folded pattern, which is in good agreement with our experimental observations shown in Figures 1 and 2. In the work, we introduce W and Dm as feature sizes characterizing the folding morphology (Supporting Information Figure S5). The higher the folding degree, the smaller the width, W, and
the higher the maximum depth, Dm. So it is easy to understand that increase of the tensile prestrain of the substrate will lead to a higher folding state, which then leads to the lower W and the higher Dm. While at the same tensile prestrain of the substrate, the higher thickness of the film will lead to both the higher W and Dm, indicating different tunability of the two parameters, Δε and hf. It should be mentioned that, too low of Dm means very weak interaction between the film and the substrate, so the interface detachments might be unavoidable in this case (e.g., thin GO layers in Figure 2a or thick but winkling GO layers in Figure 1a, in both cases, Dm < 20 μm). As concluded by Chen et al. in a recently published review,36 the creation of 3D structures from 2D building blocks enables different chemical, mechanical or physical functionalities that 6846
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
Article
ACS Nano
Figure 3. Reversible bidirectional bending properties of the highly folded-GO/latex bilayer actuators. (a) Schematic illustration of the mechanical self-assembly of the bilayer actuators. (b) Digital images of the reversible bending and recovering process of the bilayer actuator (εf = 67%, 10 mm × 10 mm) in response to n-hexane. (c) Mismatched tensile strain analysis for bending directions. The residue tensile strain (εres‑s) of the substrate on the GO-covered side, the swelling-induced tensile strain (εs‑wet) and the residue tensile strain (εs‑dry) of the substrate on the uncovered side are labeled in (a) at necessary positions, whose deviation shown in (c) is indicative of positive and negative bending directions, and generally in proportion to the bending curvatures.
On swelling in solvent, Δε = εs‑wet − εres‑s > 0, and on drying in air, Δε = εs‑dry − εres‑s < 0 (Figure 3c). Here εs‑wet and εs‑dry are the tensile strains of the uncovered side of the latex on swelling or drying, while εres‑s is that of the GO-covered side, as indicated in Figure 3a. εs‑dry is derived from the curvature of the actuator in air and εs‑wet is a sum of εs‑dry and the swelling degree (25%) of the latex rubber in n-hexane (Supporting Information Figure S8). The mismatched strain of the bilayer actuator is indicative of its bending direction. As shown in Figure 3c, the larger the residual tensile strain, the more negative mismatched strain (red gap) can be induced. So the negative bending curvature of the actuator increases with the increase of the residual tensile strains (Supporting Information Figure S6a and Table S1). On swelling, there is a net swelling-induced tensile strain loaded on the uncovered side of the latex. So εs‑wet will increase beyond the residual strain, leading to a positive bending with a positive mismatched strain (blue gap). However, we found that the positive bending curvature is not linear to the calculated mismatched strain shown in Figure 3c. One typical exclusion is the wrinkled sample prepared with 22% prestrain that does not give a high positive curvature, while both the noncompressed and highly compressed samples could offer very high positive curvatures. We think that the buckling morphology plays an important role. As observed from the detached interface, the wrinkled GO layer might have much higher elastic compressive energy than that of the folded GO layer, so it will stretch with the swelling of the substrate and thus lower the mismatched strain. Similar elongation of wrinkling polymer substrate during the swelling process has been demonstrated.38 By contrast, the highly folded GO layer was folded deeply into the substrate and
cannot be realized with planar thin films or in bulk materials. Since the residue stresses are opposite on two sides of the bilayer system, i.e., compressive in GO and tensile in the latex, as shown in Figure 3a and Supporting Information Figure S6a, once the bilayer material is cut into square slices, they will naturally undergo sheet-to-tube transitions so as to partially release the residue stresses stored in the system. GO is on the outside of the mechanical self-assembled tubes. Bigger bending angles are favorable for the higher releasing of the stresses, so the bending angles increase dramatically with increasing of the tensile prestrains. Impressively, the bilayer actuator can also bend backward in response to some organic solvents, and could recover to its initial bending state after drying in air (Figure 3b, Supporting Information Movie S1). To make it clear here we name the bending in air as negative (−) bending and the swelling-induced bending as positive (+) bending. Both are dynamically curling and highly observable. The actuator with noncompressive GO surface cracks into pieces after once immersion n-hexane, while the others with highly compressive surfaces could withstand more than 50 immersion cycles (Supporting Information Figure S6b). The absolute bending angle (or curvature, Supporting Information Figure S7) from negative bending to positive bending can be approaching to 1500° (or 3.0 mm−1). From the perspective of geometry, the value of the bending curvature, C, is proportional to the mismatched tensile strains, Δε, and inversely proportional to the total thickness of the actuator, t, namely,37 C ∝ Δε /t
(2) 6847
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
Article
ACS Nano
Figure 4. Solvent-responsive properties of the highly folded-GO/latex bilayer actuators. (a) Actuation performance of a bilayer actuator (εf = 59%) in response to various solvents. (b) THF-responsive actuation performance of the bilayer actuators with various compressive strains. (c) Forward and (d) backward actuation performance comparisons between our shape-changeable actuator (εf = 67%) and recently reported other actuators. An ultralarge and reversible deformation was observed at a relative short response time. All the above actuators have the same size, 10 mm × 10 mm.
the actuator could bend positively as compared to that with the noncompressed GO layer. To explore the potential actuation performances of the actuators, a bilayer actuator (10 mm × 10 mm, εf = 59%) was studied in response to various different solvents and the results are shown in Figure 4a. The actuator responds faster in THF and benzene than in n-hexane and dichloromethane (DCM) (Supporting Information Movie S2). This is in good agreement with the swelling rate of pure latex substrate in these solvents (Supporting Information Figure S8). During deswelling in air, however, the actuator recovers faster with THF and DCM than with n-hexane and benzene, which is generally related to the volatilizing rates of various solvents and their Van der Waals forces with the latex substrate. The backward recovery process is experimentally 1−2 times longer than the forward response process depending on the volatilizing property of the solvents. Since THF is the solvent that induces both the fast response and the fast recovery processes, the actuation properties of the actuators in response to THF were comparatively studied with the variation of the compressive strains, as shown in Figure 4b. All these actuators bend forward but the maximum curvatures are not linear to the compressive strains. The actuator with εf = 11% exhibits the lowest responding rate and the lowest curvature. The highest responding rate and the largest curvature are achieved for the actuator with εf = 67% (Supporting Information Movie S3). Even though the traditional noncompressive GO/latex bilayer actuator is also easy to curve forward, the backward recovery is the slowest and the actuating performance is not repeatable; by contrast, the ultralarge curvatures of the 3D-shrinking-fabricated highly folded-GO/latex actuators are reversible. This is because the highly folded GO layer have a much higher surface area than the substrate (GI = 8.9)
which can totally withstand the swelling-induced area change of the substrate. As a typical performance for a typical actuator (10 mm × 10 mm, εf = 67%) in response to THF, a maximum bending angle more than 1456° (i.e., a maximum curvature of 2.75 mm−1), and an average bending rate of 78°/s could be obtained (Supporting Information Figure S9). The recovery time to the original curvature is ∼41 s, double of the response time, ∼18 s. As shown in Figure 4c,d and Supporting Information Table S3, the overall actuation performance (maximum curvature × actuator thickness) of our shape-changeable actuator are superior to most of the recently reported other actuators, with both the best forward and the best backward performances at relatively short times. Compared with the patterns fabricated through 2D shrinking approaches, the cortex-like pattern prepared through 3D shrinking has a much more complicated morphology, whose compressive strain is ideally isotropic in every radial direction, as shown in Supporting Information Figure S10. A truly simultaneous 3D shrinking can be realized by discharging the air slowly, which promotes a much higher interaction between the surface layer and the substrate and reduces the detachments at the interfaces, unlike the frequently observed detachments in 2D methods.36 For a square-sized actuator with a length/width ratio of 1.0, the bending possibilities are equivalent in either of the X and Y directions. But if the length/width ratio is increased, the bending possibilities will increase remarkably in the longer direction, as shown in Supporting Information Figure S11. In the range of 2 to 10, the actuator always bends into a tubule along the longer direction; when the ratio increases up to 15, a helical configuration will be assembled and the number of helixes increases with the increase of the length/ width ratio. This information is helpful for designing various actuators with specific functionalities. 6848
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
Article
ACS Nano
Figure 5. Typical application cases of the highly folded-GO/latex bilayer actuators. (a) A bilayer actuator used as a mechanical hand grasping cylindrical Al foil in n-hexane. (b) Schematic illustration of the bilayer actuator and maximum-lifting weight of the bilayer actuators with various compressive strain of the GO film. Digital images of the bending and recover process of (c) a worm-like swimmer, (d) a smart package, (e) two bionic flowers, and (f) a bionic mimosa. All scale bars in the images are 1 cm.
Some typical applications of the actuators were designed and demonstrated and their performances are illustrated in Figure 5. As shown in Figure 5a, the cylindrical aluminum (Al) foil can be wrapped and lifted by the bilayer actuator in n-hexane (Supporting Information Movie S4). The high roughness of the GO surface and the bending pressure induced by the swelling of the substrate are advantageous for improving the friction force. A force analysis was carried out on the cylindrical Al foil during the transferring process (inset in Figure 5b), indicating that an actuator (20 mm × 10 mm, 31.8 mg, εf = 67%) can seize and bring up a cylindrical Al foil with a mass over 1000 mg (Figure 5b). By properly cutting the materials into 1D strips, 2D crosses, and 3D biostructures, smart and fast responding actuators, such as a swimming worm (Figure 5c, Supporting Information Movie S5), a smart package (Figure 5d), two bionic flowers (Figure 5e, Supporting Information Movie S6), and a bionic mimosa (Figure 5f, Supporting Information Movie S7), have been successfully demonstrated.
It is worth mentioning that noncompressed or buckyball-like slightly compressed (εf = 11%) GO surfaces exhibit hydrophilic property with water contact angle below 90°, as shown in Figure 6a; Comparatively, the highly folded GO surfaces are hydrophobic with much higher contact angles, and the contact angel increases with the increase of the folding state. In addition, if the GO surfaces are treated with (heptadecafluoro-1,1,2,2-tetrahydrodecyl) trichlorosilane vapor, Si−Cl groups of the gas will form Si−O bonds with O−H groups of the GO surfaces (Supporting Information Figure S12), and all the contact angles are increased remarkably. An average water contact angle over 162°, i.e., superhydrophobicity, can be achieved (hf= 2 μm, εf = 67%). In addition, as shown in Figure 6b, the highly folded GO surfaces after 30 min treatment exhibit excellent superoleophilic properties so that the crude oil could spread out on the surface within 40 s, which widens its applications in the field of surface chemistry. Since the superhydrophobic and superoleophilic, microwrinkled reduced GO patterns constrained on a static PDMS 6849
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
Article
ACS Nano
Figure 6. Superhydrophobic and superoleophilic properties of the highly folded GO surface and the performance of a crude-oil collector. (a) Water contact angles of the highly folded GO surface with variation of the compressive strains. The surface was treated with (heptadecafluoro-1,1,2,2-tetrahydrodecyl) trichlorosilane vapor with a duration of 0, 5, and 30 min. Contact angles were measured by sticking the latex side to a planar glass. (b) Time-dependent contact angles of the crude oil on the modified GO surface with variation of the compressive strains. (c) Schematic illustration of a crude-oil collector aided by the superhydrophobic and superoleophilic modified GO surface and the solvent-responsive bilayer system. (d) Digital images of the collecting process of an actuator. Scale bar: 1 cm. (e) Collected oil mass ratios to the actuators with variation of the compressive strains.
the whole weight of the actuator or more than 80 times the weight of the GO itself (Supporting Information Figure S13). This finding might open up a road to the cleanup of crude-oil spills using this solvent-responsive actuator.
substrate have been demonstrated as a highly portable and recyclable oil sorbent,39 the superhydrophobic and superoleophilic, highly folded modified GO patterns constrained on the solventresponsive latex substrate are more promising, in that they are dynamic with the solvent-induced oil-collecting behaviors. As shown in Figure 6c, it is supposed that the actuator is able to bend positively upon wetting in the crude-oil spills, responsive to the various hydrocarbons present in the oil, during which the actuator assemblies into a tube toward the GO surface and the oil is pushed into the tube. Water is excluded and separated. Experimental results shown in Figure 6d,e and Supporting Information Movie S8 realized the design to a great extent. The collecting capability of the actuators increases with increasing the compressive strains and reaches highest at εf = 59%. The maximum weight of the collected oil is over 8 times
CONCLUSIONS In summary, using 2D GO as the building blocks for higher dimensional patterning, we propose and demonstrate a simple 3D-shrinking method to generate multiscale highly convoluted GO patterns on a latex balloon substrate. Gyrification in the human brain is driven by the compressive stress due to the tangential expansion of the cortical layer, while here it is driven by the compressive stress due to the tangential shrinkage of the soft substrate. The GO topography is similar to the cerebral cortex at the microscale level. The surface can be modified from 6850
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
Article
ACS Nano hydrophilic to hydrophobic or superhydrophobic and superoleophilic. More interestingly, the bilayer actuators have excellent bidirectional bending property both having large curvatures. In air, it assembles into tubes with GO layer outside; in some organic solvents, such as THF and n-hexane, it curves outward into tubes with rubber layer outside. This is a combination of residue compressive stress stored in the GO layer and the swelling-induced expansion stress in the latex. An ultralarge and reversible curvature, 2.75 mm−1, was realized from the original negative bending to the final positive bending. So the method demonstrated here is not only able to fabricate highly folded cortex-like patterns with superhydrophobic and superoleophilic properties, but also to fabricate interesting reversible bilayer actuators with ultralarge deformations for versatile applications.
Movie of the actuators responsive to THF (AVI, play speed 1× ) (AVI) Movie of a mechanical hand (AVI, play speed 1× ) (AVI) Movie of a swimming worm (AVI, play speed 2× ) (AVI) Movie of the two bionic flowers (AVI, play speed 2× ) (AVI) Movie of a bionic mimosa (AVI, play speed 2× ) (AVI) Movie of an oil collector (AVI, play speed 2× ) (AVI)
AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. ORCID
Zengyong Chu: 0000-0001-7430-6027 Author Contributions
METHODS
Z.Y.C. and Y.T. conceived and designed the experiments. Y.T. carried out all the experimental fabrications and characterizations. All the authors contributed to important discussions regarding the research. Y.T. and Z.Y.C. wrote the original paper. All the authors took part in the rewriting of the manuscript and approved the final version.
Fabrication of the Highly Folded-GO/Latex Bilayer System. The latex balloon was purchased from Xiongxian Pengshuai Latex Products Co., Ltd. (Supporting Information, Figure S14) and cleaned with ethanol. GO suspensions with various concentrations were prepared according to the modified Hummers method.40 The aqueous GO suspension is rotation-dip-coated and air-dried on the surface of an air-charged latex balloon, forming a thin and uniform GO coating. Highly folded-GO surfaces can be induced by discharging the air slowly. The thickness of the GO film was controlled by varying the concentrations of the GO suspension, from 2.0 mg/mL to 7.0 mg/mL. The tensile prestrains of the latex substrate were mainly controlled by varying the diameters of the air-changed balloon, from 5.0 cm to ∼24.0 cm. Characterizations. A JSM-6700F microscope was used to characterize the micro/nano-structures of the folded surfaces. The surface was spurted with a thin film of Au for SEM imaging. A JEM2100F electron microscope was used to record transmission electron microscopy (TEM) images. Atomic force microscopy (AFM) was performed using a Veeco DI Nanoscope Multi Mode V system. Fourier transform infrared spectroscopy (FT-IR) spectra were obtained by a Bruker TENSOR-27 Fourier transform infrared spectrophotometer. X-ray photoelectron spectroscopy (XPS) was carried on a Kα 1063 instrument with focused monochromatized Al Kα radiation. The static water contact angle was measured using JC2000 × 1 contact angle and surface tension analysis system with water droplet size of 10 μL. Actuation Performance. The bilayer films were cut into pieces with different length × width sizes (typically 10 mm × 10 mm). The mechanically self-assembled tubes were immersed in various common solvents and their actuation performance was recorded using a video camera recorder. The solvents were purchased in chemical grade and used as received. The crude oil with a viscosity of 4000 mPa·s was provided by China University of Petroleum and used after mixing with 25 vol% of gasoline. The bending angle or curvature was collected by analyzing the real-time geometry changes of the actuator.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS The project was financially supported by Hunan Provincial Natural Science Foundation for Distinguished Young Scholars (No. 14JJ1001) and National Natural Science Foundation of China (No. 61574172). Thanks to Dr. Duo Zhang for his helpful mechanical discussions. Thanks to Mr. Xianshuai Zou, Dali Xu, and Miss Binru Xu for their assistance in the experiments. REFERENCES (1) Assender, H.; Bliznyuk, V.; Porfyrakis, K. How Surface Topography Relates to Materials’ Properties. Science 2002, 297, 973−976. (2) Xu, S.; Yan, Z.; Jang, K.-I.; Huang, W.; Fu, H.; Kim, J.; Wei, Z.; Flavin, M.; McCracken, J.; Wang, R.; Badea, A.; Liu, Y.; Xiao, D.; Zhou, G.; Lee, J.; Chung, H. U.; Cheng, H.; Ren, W.; Banks, A.; Li, X.; et al. Assembly of Micro/Nanomaterials into Complex, ThreeDimensional Architectures by Compressive Buckling. Science 2015, 347, 154−159. (3) Muller, M.; Karg, M.; Fortini, A.; Hellweg, T.; Fery, A. WrinkleAssisted Linear Assembly of Hard-Core/Soft-Shell Particles: Impact of the Soft Shell on the Local Structure. Nanoscale 2012, 4, 2491−2499. (4) Genzer, J.; Groenewold, J. Soft Matter with Hard Skin: From Skin Wrinkles to Templating and Material Characterization. Soft Matter 2006, 2, 310−323. (5) Goel, P.; Kumar, S.; Sarkar, J.; Singh, J. P. Mechanical Strain Induced Tunable Anisotropic Wetting on Buckled PDMS Silver Nanorods Arrays. ACS Appl. Mater. Interfaces 2015, 7, 8419−8426. (6) Jin, C.; Khare, K.; Vajpayee, S.; Yang, S.; Jagota, A.; Hui, C. Y. Adhesive Contact between a Rippled Elastic Surface and a Rigid Spherical Indenter: from Partial to Full Contact. Soft Matter 2011, 7, 10728−10736. (7) Zhang, L.; Lang, X.; Hirata, A.; Chen, M. Wrinkled Nanoporous Gold Films with Ultrahigh Surface-Enhanced Raman Scattering Enhancement. ACS Nano 2011, 5, 4407−4413. (8) Sun, Y.; Choi, W. M.; Jiang, H.; Huang, Y. Y.; Rogers, J. A. Controlled Buckling of Semiconductor Nanoribbons for Stretchable Electronics. Nat. Nanotechnol. 2006, 1, 201−207. (9) White, M. S.; Kaltenbrunner, M.; Głowacki, E. D.; Gutnichenko, K.; Kettlgruber, G.; Graz, I.; Aazou, S.; Ulbricht, C.; Egbe, D. A. M.;
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.7b01937. Characteristic properties of raw materials, detailed processing conditions and various parameters of the folding morphology, fabrication comparison between this work and recently reported other methods, performance comparison between our actuator and recently reported other actuators (PDF) Movie of the bending cycle life of an actuator (AVI, play speed 2× ) (AVI) Movie of the Actuators responsive to different solvents (AVI, play speed 1 × ) (AVI) 6851
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
Article
ACS Nano
isopropylacrylamide) Nanocomposite Hydrogels. J. Mater. Chem. A 2014, 2, 15633−15639. (31) Ji, M.; Jiang, N.; Chang, J.; Sun, J. Near-Infrared Light-Driven, Highly Efficient Bilayer Actuators Based on Polydopamine-Modified Reduced Graphene Oxide. Adv. Funct. Mater. 2014, 24, 5412−5419. (32) Yang, K.; Feng, L.; Liu, Z. Stimuli Responsive Drug Delivery Systems Based on Nano-Graphene for Cancer Therapy. Adv. Drug Delivery Rev. 2016, 105, 228−241. (33) Stuart, M. A. C.; Huck, W. T. S.; Genzer, J.; Muller, M.; Ober, C.; Stamm, M.; Sukhorukov, G. B.; Szleifer, I.; Tsukruk, V. V.; Urban, M.; Winnik, F.; Zauscher, S.; Luzinov, I.; Minko, S. Emerging Applications of Stimuli-Responsive Polymer Materials. Nat. Mater. 2010, 9, 101−113. (34) Li, C.; Zhang, X.; Cao, Z. Triangular and Fibonacci Number Patterns Driven by Stress on Core/Shell Microstructures. Science 2005, 309, 909−911. (35) Cao, G.; Chen, X.; Li, C.; Ji, A.; Cao, Z. Self-Assembled Triangular and Labyrinth Buckling Patterns of Thin Films on Spherical Substrates. Phys. Rev. Lett. 2008, 100, 036102. (36) Chen, P. Y.; Liu, M. C.; Wang, Z. Y.; Hurt, R. H.; Wong, I. Y. From Flatland to Spaceland: Higher Dimensional Patterning with Two-Dimensional Materials. Adv. Mater. 2017, 1605096. (37) Timoshenko, S. Analysis of Bi-Metal Thermostats. J. Opt. Soc. Am. 1925, 11, 233−255. (38) Li, B.; Du, T.; Yu, B.; van der Gucht, J.; Zhou, F. CaterpillarInspired Design and Fabrication of a Self-Walking Actuator with Anisotropy, Gradient, and Instant Response. Small 2015, 11, 3494− 3501. (39) Feng, C.; Yi, Z.; She, F.; Gao, W.; Peng, Z.; Garvey, C. J.; Dumée, L. F.; Kong, L. X. Superhydrophobic and Superoleophilic Micro-Wrinkled Reduced GO as a Highly Portable and Recyclable Oil Sorbent. ACS Appl. Mater. Interfaces 2016, 8, 9977−9985. (40) Hummers, W. S.; Offeman, R. E. Preparation of Graphitic Oxide. J. Am. Chem. Soc. 1958, 80, 1339−1339.
Miron, M. C.; Major, Z.; Scharber, M. C.; Sekitani, T.; Someya, T.; Bauer, S.; Sariciftci, N. S. Ultrathin, Highly Flexible and Stretchable PLEDs. Nat. Photonics 2013, 7, 811−816. (10) Chung, J. Y.; Nolte, A. J.; Stafford, C. M. Surface Wrinkling: A Versatile Platform for Measuring Thin-Film Properties. Adv. Mater. 2011, 23, 349−368. (11) Chung, J. Y.; Chastek, T. Q.; Fasolka, M. J.; Ro, H. W.; Stafford, C. M. Quantifying Residual Stress in Nanoscale Thin Polymer Films via Surface Wrinkling. ACS Nano 2009, 3, 844−852. (12) Wang, Z.; Tonderys, D.; Leggett, S. E.; Williams, E. K.; Kiani, M. T.; Spitz Steinberg, R.; Qiu, Y.; Wong, I. Y.; Hurt, R. H. Wrinkled, Wavelength-Tunable Graphene-Based Surface Topographies for Directing Cell Alignment and Morphology. Carbon 2016, 97, 14−24. (13) Bowden, N.; Brittain, S.; Evans, A. G.; Hutchinson, J. W.; Whitesides, G. M. Spontaneous Formation of Ordered Structures in Thin Films of Metals Supported on an Elastomeric Polymer. Nature 1998, 393, 146−149. (14) Tan, Y. L.; Jiang, Z. H.; Chu, Z. Y. Surface Wrinkles Based on Polymer Substrate: Biomimetic Construction, Micro-Regulation and Applications. Acta Polym. Sin. 2016, 1508−1521. (15) Chen, X.; Yin, J. Buckling Patterns of Thin Films on Curved Compliant Substrates with Applications to Morphogenesis and ThreeDimensional Micro-Fabrication. Soft Matter 2010, 6, 5667−5680. (16) Cao, C.; Chan, H. F.; Zang, J.; Leong, K. W.; Zhao, X. Harnessing Localized Ridges for High-Aspect-Ratio Hierarchical Patterns with Dynamic Tunability and Multifunctionality. Adv. Mater. 2014, 26, 1763−1770. (17) Chen, P. Y.; Sodhi, J.; Qiu, Y.; Valentin, T. M.; Steinberg, R. S.; Wang, Z.; Hurt, R. H.; Wong, I. Y. Multiscale Graphene Topographies Programmed by Sequential Mechanical Deformation. Adv. Mater. 2016, 28, 3564−3571. (18) Jin, H. C.; Abelson, J. R.; Erhardt, M. K.; Nuzzo, R. G. Soft Lithographic Fabrication of an Image Sensor Array on a Curved Substrate. J. Vac. Sci. Technol., B: Microelectron. Process. Phenom. 2004, 22, 2548−2551. (19) Shin, G.; Jung, I.; Malyarchuk, V.; Song, J.; Wang, S.; Ko, H. C.; Huang, Y.; Ha, J. S.; Rogers, J. A. Micromechanics and Advanced Designs for Curved Photodetector Arrays in Hemispherical ElectronicEye Cameras. Small 2010, 6, 851−856. (20) Creighton, M. A.; Zhu, W.; van Krieken, F.; Petteruti, R. A.; Gao, H.; Hurt, R. H. Three-Dimensional Graphene-Based Microbarriers for Controlling Release and Reactivity in Colloidal Liquid Phases. ACS Nano 2016, 10, 2268−2276. (21) Wang, Z.; Lv, X.; Chen, Y.; Liu, D.; Xu, X.; Palmore, G. T. R.; Hurt, R. H. Crumpled Graphene Nanoreactors. Nanoscale 2015, 7, 10267−10278. (22) Hallett, M. B.; von Ruhland, C. J.; Dewitt, S. Chemotaxis and the Cell Surface-Area Problem. Nat. Rev. Mol. Cell Biol. 2008, 9, 662− 662. (23) Tallinen, T.; Chung, J. Y.; Biggins, J. S.; Mahadevan, L. Gyrification from Constrained Cortical Expansion. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 12667−12672. (24) Beloussov, L. V.; Troshina, T. G.; Mansurov, A. N. Mechanical Feedback in Morphogenesis and Cell Differentiation. Biophysics 2008, 53, 575−579. (25) Striedter, G. F.; Srinivasan, S.; Monuki, E. S. Cortical Folding: When, Where, How, and Why? Annu. Rev. Neurosci. 2015, 38, 291− 307. (26) Tallinen, T.; Chung, J. Y.; Rousseau, F.; Girard, N.; Lefèvre, J.; Mahadevan, L. On the Growth and Form of Cortical Convolutions. Nat. Phys. 2016, 12, 588−593. (27) Li, B.; Jia, F.; Cao, Y. P.; Feng, X. Q.; Gao, H. Surface Wrinkling Patterns on a Core-Shell Soft Sphere. Phys. Rev. Lett. 2011, 106, 2155− 2161. (28) Ionov, L. Polymeric Actuators. Langmuir 2015, 31, 5015−5024. (29) Stoychev, G.; Ionov, L. Actuating Fibers: Design and Applications. ACS Appl. Mater. Interfaces 2016, 8, 24281−24294. (30) Zhang, E.; Wang, T.; Hong, W.; Sun, W.; Liu, X.; Tong, Z. Infrared-Driving Actuation Based on Bilayer Graphene Oxide-Poly(N6852
DOI: 10.1021/acsnano.7b01937 ACS Nano 2017, 11, 6843−6852
