Solid−Liquid Self-Adaptive Polymeric CompositePei Dong,†,⊥ Alin Cristian Chipara,†,⊥ Phillip Loya,† Yingchao Yang,† Liehui Ge,† Sidong Lei,† Bo Li,†

Gustavo Brunetto,‡ Leonardo D. Machado,‡ Liang Hong,† Qizhong Wang,† Bilan Yang,† Hua Guo,†

Emilie Ringe,† Douglas S. Galvao,‡ Robert Vajtai,† Mircea Chipara,§ Ming Tang,† Jun Lou,*,†

and Pulickel M. Ajayan*,†

†Department of Materials Science and NanoEngineering, Rice University, 6100 Main Street, Houston, Texas 77005, United States‡Applied Physics Department, State University of Campinas, Campinas-SP 13083-959, Brazil§Department of Physics and Geology, University of TexasPan American, 1201 West University Drive, Edinburg, Texas 78539,United States

*S Supporting Information

ABSTRACT: A solid−liquid self-adaptive composite (SAC) is synthesized using a simple mixing−evaporation protocol, withpoly(dimethylsiloxane) (PDMS) and poly(vinylidene fluoride) (PVDF) as active constituents. SAC exists as a porous solidcontaining a near equivalent distribution of the solid (PVDF)−liquid (PDMS) phases, with the liquid encapsulated and stabilizedwithin a continuous solid network percolating throughout the structure. The pores, liquid, and solid phases form a complexhierarchical structure, which offers both mechanical robustness and a significant structural adaptability under external forces. SACexhibits attractive self-healing properties during tension, and demonstrates reversible self-stiffening properties under compressionwith a maximum of 7-fold increase seen in the storage modulus. In a comparison to existing self-healing and self-stiffeningmaterials, SAC offers distinct advantages in the ease of fabrication, high achievable storage modulus, and reversibility. Suchmaterials could provide a new class of adaptive materials system with multifunctionality, tunability, and scale-up potentials.

KEYWORDS: hierarchical structure, self-adaptive, self-healing, self-stiffening, solid−liquid composite

1. INTRODUCTION

Many natural materials have adaptive capabilities to activelyrespond to the applied external loads, i.e., Polypterus senegalus.1

When this “living fossil” fish is challenged by external force, itsreinforcing composite can protect itself from a biting attack,due to its unique deformation mechanisms. In general, the“adaptiveness” of the inner structures allows structural andcorresponding property changes under external driving forces,and these changes can often recover to their original stateswhen the stimulation is removed. Design of artificial self-adaptive material should focus on not only changing propertiesof the materials, but also designing new morphologies andphase distribution schemes to increase adaptability of the innerstructure.2 Creating a solid−liquid material could be anattractive new direction to realize the ability to reorganizeand adapt the inner structure to the external load, leading toexciting self-adaptive behaviors including self-healing andreversible self-stiffening.As an important class of self-healing materials, the recently

reported microencapsulated,3 microvascular,4 and vascular5

material has attracted much attention owing to its solid−liquidphase distributions.3 Generally, the liquid phase is encapsulatedby a solid shell in these materials. After load-inducedmicrocrack formation and its penetration into the solid shell,the trapped liquid flows out to fill the cracks. Such structuresoffer the material the ability to autonomously heal the cracks3,6

and restore properties such as mechanical strength andelectrical conductivity.7 Unfortunately, this kind of self-healingis not repeatable, due to limited fluidity that arises from the factthat the liquid resides inside the solid shell instead ofpercolating throughout the whole solid matrix. Additionally,the complicated fabrication process8 presents a large challengefor practical utilization in many potential applications.With biological processes as inspiration, such as bone

remodeling,9 muscle development,10 and blood vessel structureevolution,11 several artificial self-stiffening material systems have

Received: November 6, 2015Accepted: December 31, 2015Published: December 31, 2015

Research Article

www.acsami.org

© 2015 American Chemical Society 2142 DOI: 10.1021/acsami.5b10667ACS Appl. Mater. Interfaces 2016, 8, 2142−2147

recently been developed,12,13 despite the design complexities insuch systems. The increase of the storage modulus, a size-independent parameter, is used to characterize the self-stiffening behavior.14 Previously, Carey et al. reported 12%self-stiffening by using the carbon nanotubes/PDMS softcomposite under DMA testing at a 5% strain amplitude andat a frequency of 5 Hz.12 The storage modulus was less than 17MPa at 100 °C. Although no storage modulus data waspresented, Agrawal et al. also reported a 90% increase ofstiffness (defined as the force/displacement ratio) in liquidcrystal elastomers, when subjected to an amplitude of 5% strainand a frequency of 5 Hz.13 However, these all-solid systemshave limited ability to accommodate structural changes and lackcrucial reversibility.In retrospect, it would seem that, to make a highly adaptive

material, the best option would be to create a reconfigurableinternal structure. This could be achieved if one were to createa solid−liquid continuum framework at the appropriate scalesso that the liquid could deform and dynamically respond withinthe deformable skin. To prepare such a composite, it would bepossible to choose one solid polymer and a viscous liquidpolymer as constituents. Inherent porosity (with trapped air)could also enable this structure to dynamically accessmorphologically varying configurations with limited swellingor other local mechanical damage such as internal fracture.Guided by this design principle, poly(vinylidene fluoride)(PVDF) was chosen as the solid skeleton for the designedcomposite in the current study, due to its good thermal andmechanical properties.15−18 Polydimethylsiloxane (PDMS), aviscous liquid, was used as the soft liquid phase. To maintainthe liquid phase of the PDMS, no curing agent was used duringthe fabrication process. Tetrahydrofuran (THF) was used tofully dissolve PVDF and disperse PDMS. Previously, a solventevaporation method was used as a facile method to encapsulatethe self-healing agents.19 Here, the probe sonication andmagnetic stirring were used simultaneously to ensure completemixing of all the components. The drying process not onlyremoves the solvent and forms a porous structure, but alsocauses the PVDF and PDMS to demix into an interpenetratingsolid skeleton and liquid phase. This creates a highly complexinterencapsulated structure, which provides structural integrityalthough nearly 50 wt % of the composite is liquid. Bycontrolling the synthesis temperature, it is possible to retainmost of the liquid PDMS phase while evaporating the majorityof the THF solvent. The distributed solid phase PVDF,encapsulating the liquid phase PDMS, provides the materialwith solid mechanical robustness while maintaining viscousadaptability, allowing for exciting self-healing and self-stiffeningbehaviors in a reversible fashion upon external stimulations.

2. EXPERIMENTAL SECTIONSAC Preparation. PVDF (Alfa Aesar, MDL: MFCD00084470,

melt viscosity (230°, 100/s) 23 500−29 500 poise) and hydroxyterminated PDMS (Sigma-Aldrich, viscosity of 18 000−22 000 cSt)with the same weight were added to the THF solvent in order. Theweight percentage of PVDF and PDMS in THF is lower than 10%.These two components were mixed using probe sonication (VCX 750,Sonics & Materials, Inc.) and magnetic stirring simultaneously foraround 3 h. The mixture was sonicated until at least 50% of the solventwas evaporated and was further evaporated in a vacuum ovenovernight at 95 °C. Then, the material was taken out and cooled toroom temperature. Soft, intermediate, and hard SAC materials wereprepared by changing the probe sonication temperature, which directlyleads to different evaporation speeds. The soft SAC was sonicated at

95% amplitude without heating the probe. Intermediate and hard SACwere prepared with a probe temperature of 65 and 85 °C, respectively,at 98% amplitude.

Morphology Characterization. The optical image of the samplewas taken using a Canon T3i digital camera. The SEM images weretaken by FEI Quanta 400 ESEM, FEI Helios Nanolab 660, and ZeissAuriga Crossbeam. The FIB procedure was performed using FEIHelios Nanolab 660 and Zeiss Auriga Crossbeam. Movie S1 wasrecorded using FEI Helios Nanolab 660. The EDS mapping wasperformed using Quantax EDS from Bruker.

Mechanical Characterization. For in situ probing, a probe with∼10 μm diameter tip attached to a nanomanipulator (Zyvex S-100)was used to manipulate a bead inside a FEI Quanta 400 ESEM. TheFEI Quanta 400 ESEM recorded Movies S2 and S3. For the in situtensile testing, a Gatan microtensile tester with a 2N load cell was usedto perform the tensile test inside a FEI Quanta 400 ESEM. Aprenotched rectangular sample was clamped on sample stages beforerepeated loading cycles. The notch length is ca. 50% of the sampledimension along the longitudinal notch direction. In the tensile test,the sample was then loaded at a displacement rate of 0.5 mm/min.The FEI Quanta 400 ESEM recorded Movie S4. For the DMA testing,the samples were hand-cut with a razor blade. The mechanicalproperties of this material were tested and analyzed using a TAInstruments DMA Q800. Tests were carried out under a constantfrequency of 5 Hz with an amplitude of 30 μm (fixed displacement) byup to 400 000 cycles. The DMA tests were conducted at the roomtemperature.

3. RESULT AND DISCUSSION

The as-fabricated PVDF−PDMS self-adaptive composite(SAC), shown in Figure 1a, contains both solid and liquidphases and possesses an interesting hierarchical microstructure(Figure 1) as discussed below. The scanning electronmicroscope (SEM) image of SAC in Figure 1b shows itsporous structure as expected from the evaporation process, withmany bead-like structures stacking each other (one such bead ishighlighted by the red dashed circle). Figure 1c is the highmagnification SEM image of the shell of the bead. An algorithmwas used to analyze many SEM images, and from a sample of240 beads an average diameter of 45.2 μm was found, with astandard deviation of 21.9 μm. (See Supporting InformationFigures S1, S2, S3 and Table S1). The measured density of theporous SAC (Figure 1a) is ∼0.61 g/cm3.The schematic image of the SAC is shown in Figure 1e. The

silver color represents the shell of the beads, while the red colorshows the viscous phase outside. At the next (second) level inthis hierarchical structure, two types of microstructures areobserved in the interior of the beads (highlighted by the bluecolor in Figure 1e), shown in Figure 1d,f, and SupportingInformation Movie S1. Figure 1d shows the cross section of abead cut by the focused ion beam (FIB), exhibiting a spinodaldecomposition-like20 morphology presumably resulting fromthe demixing of immiscible components in the PVDF−PDMSpolymer blends. A second type of bead’s internal structure,shown in Figure 1f, consists of many spheres with diameter inthe range 100−200 nm dispersed in a matrix. Energy dispersiveX-ray spectroscopy (EDS) mapping (Figure 1h,i) shows thatsilicon is absent in the spheres but rich in the matrix, andfluorine (F) is more concentrated in the spheres. This resultsuggests that the spheres are a PVDF-rich phase and the matrixis a PDMS-rich viscous phase as Si only exists in PDMS and Fonly in PVDF. Notably, Figure 1i shows that a significantamount of PVDF is dispersed in PDMS even though they areimmiscible. It is likely that the high viscosity of PDMSkinetically prevents the complete crystallization of PVDF. Close

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.5b10667ACS Appl. Mater. Interfaces 2016, 8, 2142−2147

2143

examination of the spherical PVDF particles (Figure 1g) fromscanning transmission electron microscopy (STEM) revealsmany dendrites protruding from the sphere surface. As thesphere and dendrite are two common forms of crystallinePVDF precipitates in polymer blends,21−24 such morphologyindicates that the crystallization of PVDF is the dominant phasetransformation process in the second-level microstructure. TheXRD results in Supporting Information Figure S4 also provethe existence of PVDF crystallization. Indentation by amicrometer-sized tip within SEM confirms that the shell issolid and should be composed mainly of PVDF, whichencapsulates the mixed solid/liquid phases inside the beads.To examine the mechanical integrity and robustness of the

SACs, a probe with ∼10 μm diameter tip attached to ananomanipulator (Zyvex S-100) was first used to manipulate abead inside an SEM (FEI, Quanta 400). As shown in Figure2a,b, the bead underwent pronounced rolling under the appliedforce. However, the relative motion of the bead and the rest ofthe sample did not cause them to separate. Instead, the beadmerged with nearby beads upon rotating and exhibitsconsiderable resistance to fracture shown in SupportingInformation Movies S2 and S3. The spots on the bead wheresigns of rolling and merging are visible as highlighted with redarrows before and after in situ probe contact in Figure 2a,b.This is direct evidence of the flexibility of the structure of theSAC, made possible by the existence of the solid−liquidinterface. Upon retrieval of the probe, an extending string ofsmall liquid beads was drawn between the bead and probe, as

shown in Figure 2c and its inset. This confirms that liquid phaseis present on the bead surface or within the shell layer.A simple fracture test under repeated loading was next

conducted using a microtensile tester with a 2N load cell.Figure 2d shows the in situ SEM image of a prenotched(highlighted by the red dashed line) rectangular sampleclamped on the sample stages before the repeated loadingcycles. The notch length is ca. 50% of the sample dimensionalong the longitudinal notch direction. In the tensile test, thesample was then loaded at a displacement rate of 0.5 mm/min.The engineering strain is estimated to be ∼3.6% at the peakload. Figure 2e shows the in situ SEM image of the crackpropagation in the sample, marked by the red dash line, duringloading. The crack grew to a total length of ca. 75% of thesample dimension at the end of the loading stage. Interestingly,the crack started to disappear during the unloading stage, asshown in Figure 2f. Movie S4 shows healing behaviors from thein situ tensile testing. Higher magnification SEM images of thesample region where a crack evolves before, during the loadingprocess, and at the end of one loading−unloading cycle areshown in Figure 2g−i, respectively. The crack closure enabledby capillary and viscoelastic force of the liquid phase is evident.

Figure 1. Characterization of the hierarchical solid−liquid composite.(a) Image of solid−liquid SAC. (b) SEM image of the porous SAC.(c) High magnification SEM image of the shell of the bead. (d) SEMimage of the inner structure of the bead, displaying spinodal structure,after being cut by the focused ion beam (FIB). (e) Schematic image ofthe SAC. The silver color represents the shell of the beads, while thered color shows the viscous phase outside. The two types of the innerstructures (blue) of the bead are shown in parts d and f. (f) SEMimage of another type of inner structure of the bead, showingdistributed spheres with surrounding dendrite-like structures, afterbeing cut by the FIB. The thick layer outside is the sputtered Pt forFIB cutting. (g) Scanning transmission electron microscopy (STEM)image of the sphere with the surrounding dendrite structure,illustrating the crystallization of PVDF. (h and i) EDS mappingresults of Si (red) and F (yellow) elements, respectively, in the spherewith the surrounding dendrite structure.

Figure 2. Characterization of self-healing behaviors. (a−c) In situ SEMimages of a bead during probing enabled by a nanomanipulator (a),after probing (b), and after removing the probe (c). (d−f) The in situSEM images of the sample before loading (d), during loading (e), andduring unloading (f). (g−i) The higher magnification SEM images of asample region with an evolving crack before loading (g), duringloading (h), and during unloading (i). The schematic to illustrate theself-healing mechanism has been provided below (g−i). (j) The cyclicload−unload curve. The labels d, e, and f represent the moments whenthe previous SEM images were taken. The crack in parts e and h andhealing behavior in parts f and i can be observed repeatedly during thecyclic load−unload processes. (k) The load−displacement curvesbetween 1st and 2nd cycles and from different cycles (l).

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.5b10667ACS Appl. Mater. Interfaces 2016, 8, 2142−2147

2144

The schematic to illustrate the self-healing mechanism has beenprovided in Figure 2g−i as well. The silver sphere representsthe PVDF-rich shell, and the red area shows the PDMSwrapping the whole system. Before loading, the two beads werestacked on each other. During the loading, the two beads wereforced to separate. Either PDMS on the surface or coming fromthe interior of the beads provides a closure force due tocapillary and viscoelastic interaction which causes the crack toheal upon unloading. To verify whether such crack-closureindeed presents self-healing in the material, the cyclic load−unload curve was plotted with load levels corresponding toimages in Figure 3d−f clearly marked (Figure 2j). It is clear

from this loading cycle curve that the force needed to advancethe crack could always recover, suggesting the existence of self-healing behaviors. The load−displacement curves from differ-ent cycles are shown in Figure 2k,l. Despite some observablechanges in the curves, they in general showed very goodrepeatability for up to 41 cycles with no significant load-bearingcapability loss. We found that the crack propagates along thesame pathway upon loading in the cycling test, indicating thatself-healing does not fully restore the mechanical strength andthe fracture formed during the first cycle remains the weakestlocation in the sample. However, significant bonding must haveformed between the two sides of the crack upon unloadingbecause otherwise the force−displacement curves would haveexhibited much more dramatic changes between the first andsubsequent cycles, especially considering that the crack lengthamounts to about 75% of the sample dimension. The self-healing capability of SAC is further demonstrated at the

macroscopic level (see in Supporting Information Movie S5)where two pieces of SAC materials could be compressed intoone piece.In addition to its self-healing capability demonstrated above,

the SAC also shows pronounced reversible self-stiffeningbehaviors in dynamic mechanical analysis (DMA) cycliccompression tests (Figure 3). The dynamical mechanicalproperties were measured and analyzed using a TA InstrumentsDMA Q800 under a constant frequency of 5 Hz with amplitudeof 30 μm (fixed displacement) up to 400 000 cycles. Theincreasing of the storage modulus, as a size-independentparameter, was used to characterize the self-stiffeningbehavior.14 The maximum increase of SAC’s storage modulusis up to 683% with the maximum storage modulus achieving412 MPa (Figure 3f,g). It is important to note that themorphology and mechanical properties of the SACs can beeasily tuned by changing the synthesis parameters. In Figure 3,three different samples, of varying storage moduli (A, soft; B,intermediate; C, hard), are chosen to demonstrate theubiquitous self-stiffening behaviors in SACs. The size of theresulted beads in their hierarchical structure is affected by theevaporation rate of THF: faster solvent evaporation rate causesthe viscosity of PVDF/PDMS blend to increase more rapidly,which reduces the ability of droplets to coalesce and results insmaller bead sizes, as shown in Supporting Information TableS2. This gives us the ability to tune the storage modulus andself-stiffening behaviors of the materials (Figure 3). The SEMimages of the as-fabricated SAC samples are shown in Figure3a−c, respectively. The storage modulus evolution for the samesample in two separate tests is shown in Figure 3d−g. Thestorage modulus of sample A, shown in Figure 3d, increasedfrom 2.9 to 11.6 MPa after 400 000 cyclic loading. The samplewas fully relaxed before another test. The inserted image inFigure 3d shows the starting points of the storage moduli forpristine and relaxed sample A, which are quite similar. Similarself-stiffening behaviors were also observed in these twoseparate tests, demonstrating good repeatability. Figure 3g isthe histogram of the storage modulus at the beginning andending point of A, B, and C, before and after relaxation.Remarkably, the stiffened SAC samples are able to recover alarge part of their structural compliance after loading andrelaxation. Figure 3d shows that the storage modulus of Areturned from 11.6 MPa at the end of the DMA test to 3.3 MPaafter relaxation, which compared well to the original modulus of2.9 MPa. Similar relaxation in storage modulus is also shown inB and C, Figure 3e,f. These results clearly show that the self-stiffening phenomenon observed in SAC material is largelyreversible.The self-healing and self-stiffening behaviors possessed by

SAC are direct consequences of its internal porous structurethat incorporates solid, liquid phases in a continuous fashion.Our choices of the system components, i.e., a solvent (THF), aviscous liquid (PDMS), and a semicrystalline polymer (PVDF),contribute synergistically to the formation of this intriguingmicrostructure. THF helps disperse two immiscible compo-nents (PVDF and PDMS) into a liquid solution. Upon gradualremoval of THF, the following processes are thought to occurin the composite: (1) Creation of porous structure byevaporation of THF. Because of the high viscosity of PDMS,the void space in the solution left behind by vaporized THFcannot be completely filled during synthesis, which producesopen pores in the composite. In response to the presence ofinternal pores, the viscous PDMS/PVDF mixture undergoes

Figure 3. Characterization of self-stiffening behaviors by dynamicmechanical analysis (DMA). (a−c) SEM image of the soft SAC (a),the intermediate SAC (b), and the hard SAC (c). (d−f) The storagemodulus of soft SAC (d), intermediate SAC (e), and hard SAC (f)keeps increasing as the numbers of cycles increase. (g) The histogramimage of the storage modulus at the beginning and end point of thesoft SAC, intermediate SAC, and hard SAC before and after relaxation.The represented points with the same color have also been marked ind−f separately.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.5b10667ACS Appl. Mater. Interfaces 2016, 8, 2142−2147

2145

localized morphology changes and forms liquid beads to reducelocal surface energy. (2) Demixing of PDMS/PVDF solutionvia spinodal decomposition. Removal of THF decreases themutual solubility of PDMS and PVDF in the solution andshould eventually cause them to demix. Immiscible polymerblends often demix in the liquid phase through a spontaneousprocess known as “spinodal decomposition”.20 The intertwinedPDMS-rich and PVDF-rich domains observed inside the beads(Figure 1d) strongly resemble demixed liquid morphology afterspinodal decomposition. Similar spinodal structure is also seenin porous PVDF membranes (e.g., Figure 3 of ref 23 by Li andKatsouras et al.) prepared by an analogous vapor-induced phaseseparation method involving PVDF, an immiscible component,and a solvent.21−24 Because THF is removed more rapidly fromthe surface region than from the interior of the liquid throughevaporation, demixing is expected to occur earlier and fasternear the surface than in the bulk liquid. With continuousevaporation of THF and increased immiscibility between PVDFand PDMS, the PVDF-rich domains in the surface shell growby displacing PDMS liquid out of the surface layer, whicheventually results in a largely solid shell. Upon further cooling,PVDF in the mixture crystallizes; the spinodal decompositionmorphology could be either preserved by rapid crystallization,or replaced by dendritic PVDF morphology when thecrystallization rate is slower, which allows for morphologicalchanges through diffusion of PVDF and PDMS at the crystalgrowth front.25 More detailed characterization of thethermodynamic stability, phase separation, and crystallizationkinetics of the PVDF/PDMS/THF ternary system will becarried out in the future to further improve the control of theSAC microstructure.A key step in the structural evolution of SAC during

synthesis is the formation of a continuous solid PVDF shell onthe bead surface, which both encapsulates PDMS liquid andstabilizes the pore space. This surface layer provides structuralrobustness to the composite, but its thinness also gives the beadstructure significant deformability and flexibility that isinstrumental to SAC’s self-adaptiveness. As discussed above,we postulate that the solid shell structure originates from asurface-mode spinodal decomposition, followed by subsequentcrystallization and growth of PVDF domains.The self-stiffening behavior of SAC can be attributed to the

reduction of porosity as shown in Supporting InformationFigure S5 and the increased interaction among PVDF/PDMSbeads under loading. Densification of the porous compositestructure upon compression contributes to an increase innominal storage modulus, which is likely to be irreversible.Nevertheless, the majority of the stiffening is likely to originatefrom the viscoelastic interaction between the beads in thecomposite. When two neighboring beads are compressedagainst each other, their PVDF shells can have large elasticdeformation like two hollow rubber balls thanks to the smallthickness. As a result, the contact area and force between thebeads are increased, which contributes to the enhancement inthe stiffness of the sample. After the applied force is removed,the deformed shells can recover their original shapes, makingself-stiffening a reversible process. However, because of theinclusion of the highly viscous PDMS inside the beads, thedeformation of a bead as a whole exhibits viscoelasticcharacteristics. This explains why SAC displays a gradual ratherthan instantaneous change in storage modulus over many cyclesin DMA tests. In addition, insignificant self-stiffening was foundin samples tested at ≤135 °C (below the glass transition

temperature of PDMS as ∼−125 °C), which also demonstratesthe essential role of a viscous liquid phase in the self-stiffeningphenomenon.

4. CONCLUSIONSIn conclusion, a new type of composite, SAC, with a uniquecombination of liquid and solid phases, was synthesized by amixing−evaporation process, which forms a hierarchicalmicrostructure made of beads encapsulating liquid PDMSwithin thin shells of solid PVDF. The material processes theself-healing capability during the tensile test, due to thepresence of the liquid phase that can allow a liquid-like joiningbehavior of the crack surfaces. This material also undergoessignificant reversible self-stiffening under compression, with amaximum increase of storage modulus up to 683%, which ismuch larger than that reported for solid composites and othermaterials. Both self-healing and self-stiffening behaviors of SACrepresent the adaptive nature of the material due to theintermixed solid−liquid microstructure, which can be tuned byprocessing to get controllable mechanical behavior. In acomparison to that of traditional microencapsulated materials,the processing of SAC is simple, tunable, and scalable. Mostimportantly, by studying the mechanism, such solid−liquidhybrid systems represent a new direction in the design ofdynamic and adaptive materials which could have a largenumber of exciting applications, e.g., biocompatible materialsfor tissue engineering, and lightweight defect tolerant structuralmaterials, etc.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsami.5b10667.

Automated diameter measurement algorithm, the diam-eter distribution of the spheres before and after DMAtesting, XRD results, the normalized height of the SACsample before and after DMA and after relaxation (PDF)Movie S1 demonstrating two types of microstructuresinside the sphere, including both the spinodal (Figure1D) and sphere domain structures (Figure 1F), madefrom SEM images (AVI)Movie S2 demonstrating the rotating motion of the beadduring probing enabled by a nanomanipulator inside anSEM (AVI)Movie S3 demonstrating the merging motion of the beadduring probing enabled by a nanomanipulator inside anSEM (AVI)Movie S4 demonstrating the self-healing behavior duringthe tensile test enabled by a microtester inside an SEM(AVI)Movie S5 demonstrating the healing behavior bycompressing the two pieces of the SAC into one piece(AVI)

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: jlou@rice.edu.*E-mail: ajayan@rice.edu.Author ContributionsP.D. and A.C.C. carried out the material synthesis andcharacterization, and the experimental data analysis. P.L carriedout the probing experiment. Y.Y. helped perform the in situ

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.5b10667ACS Appl. Mater. Interfaces 2016, 8, 2142−2147

2146

tensile test. L.D.M. carried out simulation of the diameters ofthe sphere. L.H. drew the schematic image. Q.W. and B.Y.helped synthesize the material. M.T. explains the formation ofthe material in theory. J.L. and P.M.A. initiated the project. Allauthors participated in discussions. P.D., Y.Y., L.G, S. L, B. L.,M.T., J.L., and P.M.A. contributed to writing the manuscript.

Author Contributions⊥P.D. and A.C.C. contributed equally to this work.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThe authors would like to gratefully acknowledge the Air ForceOffice of Scientific Research (Grant FA9550-13-1-0084) forfunding this research, and the program director Dr. JoycelynHarrison for her guidance. L.H. and M.T. acknowledge supportfrom DOE Project DE-SC0014435. The authors also thank FEIand Zeiss Company for providing SEM and EDS images afterFIB cutting.

■ REFERENCES(1) Bruet, B. J. F.; Song, J. H.; Boyce, M. C.; Ortiz, C. MaterialsDesign Principles of Ancient Fish Armour. Nat. Mater. 2008, 7 (9),748−756.(2) Boncheva, M.; Whitesides, G. M. Self-healing Systems Having aDesign Stimulated by the Vertebrate Spine. Angew. Chem., Int. Ed.2003, 42 (23), 2644−2647.(3) White, S. R.; Sottos, N. R.; Geubelle, P. H.; Moore, J. S.; Kessler,M. R.; Sriram, S. R.; Brown, E. N.; Viswanathan, S. Autonomic Healingof Polymer Composites. Nature 2001, 409 (6822), 794−797.(4) Toohey, K. S.; Sottos, N. R.; Lewis, J. A.; Moore, J. S.; White, S.R. Self-healing Materials with Microvascular Networks. Nat. Mater.2007, 6 (8), 581−585.(5) White, S. R.; Moore, J. S.; Sottos, N. R.; Krull, B. P.; Cruz, W. A.S.; Gergely, R. C. R. Restoration of Large Damage Volumes inPolymers. Science 2014, 344 (6184), 620−623.(6) Cho, S. H.; Andersson, H. M.; White, S. R.; Sottos, N. R.; Braun,P. V. Polydimethylsiloxane-based Self-healing Materials. Adv. Mater.2006, 18 (8), 997−1000.(7) Blaiszik, B. J.; Kramer, S. L. B.; Grady, M. E.; McIlroy, D. A.;Moore, J. S.; Sottos, N. R.; White, S. R. Autonomic Restoration ofElectrical Conductivity. Adv. Mater. 2012, 24 (3), 398−401.(8) Therriault, D.; White, S. R.; Lewis, J. A. Chaotic Mixing in Three-dimensional Microvascular Networks Fabricated by Direct-writeassembly. Nat. Mater. 2003, 2 (4), 265−271.(9) Cowin, S. C. Bone Mechanics Handbook, 2nd ed.; CRC Press:Boca Raton, FL, 2001.(10) Goldberg, A. L.; Etlinger, J. D.; Goldspink, D. F.; Jablecki, C.Mechanism of Work-Induced Hypertrophy of Skeletal-Muscle. Med.Sci. Sports Exercise 1975, 7 (4), 248−261.(11) Nerem, R. M. Vascular Fluid-Mechanics, the Arterial-Wall, andAtherosclerosis. J. Biomech. Eng. 1992, 114 (3), 274−282.(12) Carey, B. J.; Patra, P. K.; Ci, L.; Silva, G. G.; Ajayan, P. M.Observation of Dynamic Strain Hardening in Polymer Nano-composites. ACS Nano 2011, 5 (4), 2715−2722.(13) Agrawal, A.; Chipara, A. C.; Shamoo, Y.; Patra, P. K.; Carey, B.J.; Ajayan, P. M.; Chapman, W. G.; Verduzco, R. Dynamic Self-stiffening in Liquid Crystal Elastomers. Nat. Commun. 2013, 4, 1739.(14) Meyers, M. A.; Chawla, K. K. Mechanical Behavior of Materials;Cambridge University Press: New York, NY, 2008.(15) Li, M.; Wondergem, H. J.; Spijkman, M.-J.; Asadi, K.; Katsouras,I.; Blom, P. W. M.; Leeuw, D. M. d. Revisiting the δ-phase ofPoly(vinylidene fluoride) for Solution-processed Ferroelectric ThinFilms. Nat. Mater. 2013, 12, 433−438.(16) Lovinger, A. J. Ferroelectric Polymers. Science 1983, 220 (4602),1115−1121.

(17) El Mohajir, B. E.; Heymans, N. Changes in Structural andMechanical Behaviour of PVDF with Processing and Thermomechan-ical Treatments. 1. Change in Structure. Polymer 2001, 42 (13), 5661−5667.(18) Brandrup, J.; Immergut, E. H.; Grulke, E. A. Polymer Handbook;Wiley/Interscience: New York, 1989; p 51−53.(19) Zhao, Y.; Fickert, J.; Landfester, K.; Crespy, D. Encapsulation ofSelf-Healing Agents in Polymer Nanocapsules. Small 2012, 8 (19),2954−2958.(20) Balluffi, R. W.; Allen, S.; Carter, W. C. Kinetics of Materials; JohnWiley & Sons: New Jersey, 2005.(21) Matsuyama, H.; Teramoto, M.; Nakatani, R.; Maki, T.Membrane Formation via Phase Separation Induced by Penetrationof Nonsolvent from Vapor Phase. I. Phase Diagram and Mass TransferProcess. J. Appl. Polym. Sci. 1999, 74 (1), 159−170.(22) Matsuyama, H.; Teramoto, M.; Nakatani, R.; Maki, T.Membrane Formation via Phase Separation Induced by Penetrationof Nonsolvent from Vapor Phase. II. Membrane Morphology. J. Appl.Polym. Sci. 1999, 74 (1), 171−178.(23) Li, M. Y.; Katsouras, I.; Piliego, C.; Glasser, G.; Lieberwirth, I.;Blom, P. W. M.; de Leeuw, D. M. Controlling the Microstructure ofPoly(vinylidene-fluoride) (PVDF) Thin Films for Microelectronics. J.Mater. Chem. C 2013, 1 (46), 7695−7702.(24) Peng, Y. L.; Fan, H. W.; Ge, J.; Wang, S. B.; Chen, P.; Jiang, Q.The Effects of Processing Conditions on the Surface Morphology andHydrophobicity of Polyvinylidene Fluoride Membranes Prepared viaVapor-induced Phase Separation. Appl. Surf. Sci. 2012, 263, 737−744.(25) Inaba, N.; Yamada, T.; Suzuki, S.; Hashimoto, T. MorphologyControl of Binary Polymer Mixtures by Spinodal Decomposition andCrystallization. 2. Further-Studies on Polypropylene and EthylenePropylene Random Copolymer. Macromolecules 1988, 21 (2), 407−414.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.5b10667ACS Appl. Mater. Interfaces 2016, 8, 2142−2147

2147