University of Florida Thesis Or Dissertation

RECONFIGURABLE NANOOPTICS ENABLED BY NOVEL SHAPE MEMORY POLYMERS

YIN FANG

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2016

To my family

ACKNOWLEDGMENTS

I really enjoyed the years that I spent in Florida. First, I would like to thank my advisor Dr. Peng Jiang, for offering tons of provision in my academia life. He is really my first mentor guided me to the right way on the academia road. I would also like to thank

Dr. Cutis Taylor, Dr. Thomas Angelini, Dr. Vito basile for their support in my professional life. I have learnt a lot of from them during my PhD research.

I would also to thank my committee members Dr.Zhaohui Tong, Dr. Yidder

Tseng, Dr. Kirk Ziegler for all their support and help.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 8

LIST OF FIGURES ...... 9

LIST OF ABBREVIATIONS ...... 12

ABSTRACT ...... 13

CHAPTER

1 BACKGROUND ...... 15

Shape Memory Polymer ...... 15 Photonic Crystals ...... 17

2 RECONFIGURABLE PHOTONIC CRYSTALS ENABLED BY NOVEL PRESSURE-REPOSIVE SHAPE MEMORY POLYMERS ...... 21

Introduction ...... 21 Methods ...... 22 Printing Photonic Crystal Patterns on Macroporous SMP Membranes...... 23 Sample Characterization...... 24 Results ...... 25 Preparation and Characterization of New SMPs...... 25 Pressure-Induced SM Recovery...... 27 Pressure-Induced SM Recovery Mechanisms...... 31 Young’s Moduli of the New Pressure-Responsive SMPs Determined by Nanoindentation ...... 32 AFM Surface Microstructure Characterization and Roughness Analysis ...... 34 Discussion ...... 36

3 CHROMOGENIC PHOTONIC CRYSTALS ENABLED BY NOVEL VAPOR- RESPONSIVE SHAPE MEMORY POLYMERS ...... 47

Introduction ...... 47 Methods ...... 48 Fabrication of Macroporous SMP Photonic Crystal Membranes ...... 48 Responses of Macroporous SMP Membranes Exposed to Acetone Vapors with Different Partial Pressures: ...... 49 Sample Characterization: ...... 49 Nanoindentation Tests: ...... 50 Scalar Wave Approximation Modeling: ...... 50

Results ...... 51 Conclusion ...... 59

4 DIRECT WRITING OF THREE-DIMENSIONAL MACROPOROUS PHOTONIC CRYSTALS ON PRESSURE-RESPONSIVE SHAPE MEMORY POLYMERS ...... 71

Introduction ...... 71 Methods ...... 72 Directly Printing and Hand-Writing 3D Macroscopic Photonic Crystal Patterns on Templated Macroporous SMP Membranes: ...... 73 Directly Writing 3D Photonic Crystal Micropatterns by AFM: ...... 73 Sample Characterization: ...... 74 Scalar Wave Approximation Optical Modeling: ...... 74 Micropatterning with 1 mm Diameter Spherical Tip: ...... 75 Nanopatterning with 20 µm Diameter Spherical Tip: ...... 75 Results ...... 75 Conclusion ...... 86

5 OPRICALLY BISTABLE MACROPOROUS PHOTONIC CRYSTALS ENABLED BY THERMORESPONSIVE SHAPE MEMORY POLYMERS ...... 97

Introduction ...... 97 Methods ...... 99 Templating Fabrication of Thermoresponsive Macroporous SMP Photonic Crystal Membranes: ...... 99 Heat-Induced SM Programming and Recovery: ...... 100 Sample Characterization: ...... 101 Scalar Wave Approximation Optical Modeling: ...... 101 Results ...... 102 Conclusion ...... 110

6 INSPIRED BY SCALABLE COLLOIDAL TEMPLATING TECHNOLOGY AND POROUS COATING THAT ENABLES THE FABRICATION OF BROADBAND TUNABLE AR COATING...... 119

Introduction ...... 119 Methods ...... 121 Templating Nanofabrication of Monolayer Macroporous 2D Photonic Crystal Membranes ...... 121 Macroporous SMP Shows Tunability of Antireflection Property Which Accurately Controlled by Morphology of Macropores ...... 122 Macroporous SMP Shows Tenability of Antireflection Property by Contacting Pressure ...... 122 Sample Characterization ...... 123 Results ...... 123 Fabrication of Tunable Antireflection Shape Memory Films ...... 124

Reconfigurable Antireflection 2D Photonic Crystals Enabled by Pressure- Responsive Shape Memory Polymer ...... 125 Reconfigurable Antireflection 2D Photonic Crystals Enabled by Ethanol- Responsive Shape Memory Polymer ...... 128 Conclusion ...... 130

LIST OF REFERENCES ...... 142

BIOGRAPHICAL SKETCH ...... 157

LIST OF TABLES

Table page

2-1 Roughness of 10  10 m2 AFM scan size of SMP surface...... 46

3-1 Roughness of 10  10 m2 AFM scan area of SMP sample surface...... 70

4-1 Dependence of the recovered line widths on the parameters of the nanoindentation-based direct writing process...... 95

4-2 Dependence of the recovered line width and height on the writing speed of the AFM tip...... 96

6-1 Linear profile roughness...... 140

LIST OF FIGURES

Figure page

1-1 Thermally activated SMP at 70C...... 20

2-1 Schematic illustration compares the SM effects between a traditional thermoresponsive SMP and the new pressure-responsive SMP...... 37

2-2 Structural, thermal, and mechanical properties of the new pressure- responsive SMPs...... 38

2-3 Arbitrary photonic crystal patterns printed on the new pressure-responsive SMP membranes...... 39

2-4 Difference in topography and PBGs between a macroporous SMP membrane dried out of water and ethanol...... 40

2-5 Pressure-dependent macropore recovery...... 41

2-6 Macropore recovery induced by pull-off forces. A typical indentation force- displacement curve showing approach and retraction segments and pull-off force obtained on a macroporous SMP membrane. Inset show ...... 42

2-7 Typical tensile stress versus strain curve for an ETPTA-co-PEGDA copolymer membrane with 1:3 ratios...... 43

2-8 3-D AFM images of a fingerprinted SMP sample with cross-sectional profiles. .. 44

2-9 Cross-sectional SEM image showing the transition from the disordered fingerprint valley region to the 3-D ordered fingerprint ridge region...... 45

3-1 Schematic illustration showing the SM effects of the new vapor-responsive SMP...... 61

3-2 A macroporous SMP membrane with 280 nm macropores after drying out of water...... 62

3-3 A macroporous SMP membrane with 280 nm macropores after exposing to an acetone vapor...... 63

3-4 Normal-incidence optical reflection spectra comparing a macroporous SMP membrane with 280 nm macropores dried out of water, liquid ethanol, and acetone vapor...... 64

3-5 Normal-incidence optical reflection spectra obtained from a macroporous...... 65

3-6 A macroporous SMP membrane exposed to acetone vapor above liquid acetone at different locations...... 66

3-7 Molecular structure...... 67

3-8 Typical DSC plot of a macroporous ETPTA 20-co-PEGDA 600 copolymer with 1:3 ratios...... 68

3-9 Comparison of the Young’s modulus of the ethanol-dried and water-dried macroporous ETPTA 20-co-PEGDA 600 (1:3 ratio) membranes indented with different forces...... 69

4-1 Schematic illustration showing the direct writing of microscopic 3D photonic crystal patterns (letters “U” and “F”) on a macroporous SMP membrane with collapsed macropores using an AFM tip...... 88

4-2 A green-colored, handwritten “UF” pattern on a translucent macroporous SMP membrane with collapsed 300 nm macropores...... 89

4-3 A micropattern “U” directly written on a macroporous SMP membrane...... 90

4-4 AFM image of a recovered area on the micropatterned letter “F”...... 91

4-5 Microscope image of microscopic lines written with increasing forces...... 92

4-6 AFM images of nanoscopic lines written with different AFM tip speeds...... 93

4-7 Photograph of a rubber stamp and an iridescent “A+” pattern printed on a translucent macroporous SMP membrane...... 94

5-1 PU-co-TPGDA copolymer...... 111

5-2 Schematic illustration showing the heat-induced programming and recovery steps of a thermoresponsive macroporous SMP photonic crystal membrane. 112

5-3 Photographs and SEM images showing the apparent changes in reflective colors and microstructures during heat-induced programming and recovery steps of a macroporous PU-co-TPGDA copolymer membrane ...... 113

5-4 AFM images and the corresponding height profiles (across the dashed lines) showing the changes in surface microstructures during heat-induced programming and recovery steps...... 114

5-5 Normal-incidence optical reflection spectra showing the permanent, deformed, and thermally recovered states of a macroporous PU-co-TPGDA copolymer membrane with 300 nm macropores...... 115

5-6 Time-resolved normal-incidence optical reflection spectra during thermally induced recovery of a macroporous PU-co-TPGDA copolymer membrane with collapsed 300 nm macropores...... 116

5-7 Reflection amplitudes at 525 nm wavelength and the corresponding photographs of a macroporous PU-co-TPGDA copolymer membrane ...... 117

5-8 Complete DSC plot of PU-co-TPGDA copolymer showing both heating and cooling cycles...... 118

6-1 Schematic drawing the shape memory effect at nanoporous SMP film...... 131

6-2 Arbitrary patterns printed on nanoporous SMP membrane...... 132

6-3 Various morphology between water dried area and contact printed area...... 133

6-4 Normal-incidence optical transmission spectra obtained through a nanoporous SMP film of 100nm nanopores...... 134

6-5 Nanoporous film rinsed by water and ethanol...... 135

6-6 Various morphology between water dried area and ethanol area...... 136

6-7 Normal-incidence optical transmission spectra obtained through a nanoporous SMP film of 100nm nanopores...... 137

6-8 Templated fabrication of close packed 100nm microsphere on glass substrate...... 138

6-9 Photograph of patterned structures on nanoporous SMP film...... 139

LIST OF ABBREVIATIONS

AFM Atomic Force Microscope

DSC Differential scanning calorimetry

EDS Energy dispersive spectroscopy

SEM Scanning electron spectroscopy

SAXS Small angle X-ray Scattering

TGA Thermogravimetric analysis

UV/VIS Ultraviolet-visible

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

RECONFIGURABLE NANOOPTICS ENABLED BY NOVEL SHAPE MEMORY POLYMERS

Yin Fang

August 2016

Chair: Peng Jiang Major: Chemical Engineering

Smart shape memory polymers (SMPs) have memorized functionality, which can memorize their permanent shape in response to a peripheral stimulus, such as heat, light, solvent, IR, electricity and magnetic. They have been broadly demonstrated for a variety of applications ranging from biomedical devices to aerospace morphing structures. However, most of the current SMPs are thermoresponsive and their applications are impeded by slow shape memory recovery, and tediously heat- demanding programming and recovery steps. Here, by incorporating scientific principles drawn from two incoherent fields that do not generally intersect - the self-assembly photonic crystals and SMP materials, we have developed a novel type of SMP that possesses unique "cold" programming and prompt shape recovery triggered by applying a small contact pressure, contact shear force, or exposing the samples to various organic vapors, such as acetone, ethanol, and toluene. These nano-structural shape memory materials are greatly dissimilar from current available bulk SMPs as they empower orders of magnitude faster response, striking chromogenic effects, and room- temperature actions for the entire shape memory cycles. They are promising for many applications ranging from recyclable chromogenic vapor sensors to smart nanooptical

devices. Moreover, this interdisciplinary incorporation offers a simple and sensitive optical technique for investigating the fascinating shape memory effects at nanoscale.

CHAPTER 1 BACKGROUND

Shape Memory Polymer

Shape memory materials (SMPs) are a type of smart materials. Their prominent performance endows the memory functionality on demand when the materials expose to certain stimuli. In the past decades, SMPs have gained increased attention compared to their alloy counterparts (e.g., nitinol alloy) due to their low cost, light weight, easy of synthesis, and biocompatibility.1-5 The remarkable advantages draw their practical applications in biomedical device, aerospace device, chemical and biosensor, optical detector, tunable textiles, and actuators.6-11 Through the numerous study, people recapitulate the shape memory effect of bulk shape memory materials is typically reached in a three-step procedure that includes programming, storage, and recovery.

Programming involves deforming polymers from their permanent shape (rod in Fig. 1) to a temporary shape (spiral in Fig. 1). When the sample is heated above the polymer glass transition temperature (Tg), the polymer intrinsic structure is compliant and yield to be guided by the external force. After the compliant sample has been deformed, the ambient temperature intends to decrease below Tg to freeze the motivated polymer chains and confine them in the temporary shape. The elastic entropy drives the freeze polymer chains re-active and recovers to their permanent shape.5 Owe to the unique chemical structures of shape memory polymer, the intrinsic steric structure determine the shape memory effect of the particular environment. In the classic thermally induced shape memory polymers, the entangled polymer chains are envisioned as polymer net.

The contiguous polymer net consisted of nodes and lines. The nodes are formed by the kinks of entangled polymer chains consist of the covalent bond and physical interaction

between two vicinity polymer chains named netpoints or hard segments. These formulations normally offer highest thermal transitions Tg. The polymer chains that connect the two netpoints are formed by “switch segment”, which usually consist of side chains or functional groups that are the second highest Tg called soft segments. In the programming process, the polymer switches start to relax when the ambient temperature above Tg. The relaxed switches tend to certain orientation upon loading of external stress and elongate the persistence length of polymer chains. After withdrawing the external loading and decreasing the ambient temperature below Tg, the elongated persistence length of polymer chains is kept at the current stage because of the “frozen- in” temporary confinement of polymer chains. The shape recovery is driven by entropy elasticity of the soft segments. The elongated persistence length of polymer chains recoil and reorganized to their stress-free state when the ambient temperature reheat above the Tg in accordance with the second law of thermodynamic (Boltzmann

Equation).9,12-14In the recently study, the other stimuli-responsive shape memory polymer spring up like bamboo shoots after spring rain, such as light, IR, electricity, solvent, magnetic.10,15-27As a sweeping recapitulate, the versatile triggering helps SMPs to perform more broad applications instead of classic thermal-induced shape memory polymers.28,29As illustrated by the demo in Fig 1-1, the recovery time for thermally triggered SMPs is regularly time-consuming (on the order of minutes), obstructing many applications that need the fast response. So the shape memory effect from the bulk materials to the interface is avid of accelerating the response. In addition, the various optical or electricity readout in the shape memory process also demonstrate unparalleled advantages and help people to get more accurate perspective.19,28,30-

36Compared to the typical SMPs, pressure-responsive SMPs provide an extensive degree of tailor-ability to practically accommodate application on demand (e.g., functioning at room temperature), which have received less investigation. Additionally, most of the current SMP applications relate to leveraging the SM effect on bulk materials, where the various length scale is extremely large (on the order of centimeters). However, an intriguing practical applications for all SMPs, largely unexplored, is their capability to convert shape at small scale (micrometers or nanometers).37-45

Photonic Crystals

Photonic crystals are the periodic structure in materials which repeat the refractive index from high refractive index to the low refractive index( air cavities). The specific structure assembles the refractive index of materials from 1D, 2D, 3D dimensions, and modulate the light propagation in this structure.46-48The difference of group velocity in the process of light propagation leads to the difference of energy flow to form the forbidden photonic band gap. In the electromagnetic and elastic waves, the velocity is defined by:

Vg=dw/dk (1-1) where w and k are an angular frequency and a wavevector. The diagram of w vs k normally relates to the forbidden photonic band. In the classic Brillouin zone(BZ), the light propagation speed at the edge is expressively slower(Vg~0), which improved the interaction of the photon and ban the energy flow, so called complete forbidden photonic bandgap. In the range of photonic bandgap, there is no light propagation.

Attribute-based forbidden photonic bandgap, the light propagation can be manipulated at specific frequency along the certain direction.49-52

The full-color countenance can be easily modified by fine-tuning the lattice spacing and effective refractive index in photonic crystals according to the Bragg

Diffraction.53 nλ=2neffdsinθ (1-2) where the λ is the incident wavelength, neff is effective refractive index, d is planar spacing, and θ is the incident angle.

Expecting to prosper from periodic patterning of polymers and nanoparticles, the photonic crystals fabrication by bottom-up self-assembly have been recognized as a second chance at life. Instead of conventional “top-down” lithography method, the

“bottom-up” self-assembled photonic crystals extremely decrease the cost, which draws their potential application from bi-pulsed laser to optically tunable device, such as biochemical sensor, full-color tunable display, mechanically optical sensor.48,54-

57However, the optically bistable photonic crystal based on the actively soft materials is still largely unexplored.

In nature, the inventors create tons of wonders in their long history evolution.

Inspired by the complimentary wonderfulness of the nature, scientists have more comprehending the biological masterpiece in series of their splendid work, such as firefly lanterns, moth eyes, butterfly wings, octopus skin, gecko feet, and to fabricate the microlens light emitting diodes, antireflection solar panels, toxic vapor sensor, anti- counterintuitive camouflage and tunable adhesion of micro-robot.58-66Effectively improving the properties of composite materials served mankind is always the target pursued by frontier scientist. Recently, people focus on adapting property of materials on demand. For example, the digital cameras capture the environmental view at 180-

degree field without aberration heuristic by arthropods (ants, beetles, arachnids), which have hemispherical eyes to support broad vision to avoid predators. The three layers’ structural systems of the skin of cephalopods help people to explore the artificial optoelectronic camouflage for military soldiers. The cylindrical Bragg mirror was found on the legs of rare insects, which can tremendously enrich the light extraction and significantly improve the efficiency of LED. A novel touch screen was designed, which obtains prominent current transport, unprecedented mechanical properties and unique antireflection properties inspired by miraculous spider web.58,61,65-68

Inspired by the magical skin of chameleon, our goal of this project is to develop new pressure-responsive SMPs, which integrating the fabrication of periodically structural patterns and shape memory materials.69 Accompanying by the conspicuous color changes, the intrinsic micro-or nanostructure has obvious change which would be exploited by investigating the nanoscale deformation and recovery of macroporous SMP films with 3-D ordered nanopores. In simplest, reinforcing the responsive time and sensitivity of shape memory effect from bulk materials to the interface is the inevitable development trend. Coincidentally, our new approach demonstrates a series of porous polymers that enable unusual cold programming and 1-2 orders-of-magnitude faster response rate than current SMPs.

Figure 1-1.Thermally activated SMP at 70C.1

CHAPTER 2 RECONFIGURABLE PHOTONIC CRYSTALS ENABLED BY NOVEL PRESSURE- REPOSIVE SHAPE MEMORY POLYMERS

Introduction

Shape-memory (SM) in traditional SM polymers (SMPs) is typically achieved in three steps including programming, storage and recovery. Programming involves deforming a bulk SMP sample from its permanent shape to a temporary configuration.38,44,70,71This ‘hot’ process is usually done above a specific transition temperature (Ttrans), such as the polymer glass transition temperature (Tg), to leverage the compliant properties of SMPs at high temperature. Once the sample is deformed, it is cooled below Ttrans to ‘freeze’ in the temporary shape. Recovery occurs when the sample is reheated to the vicinity of Ttrans, which increases polymer chain mobility and allows the SMP to return to its permanent shape via entropy elasticity.28,39

Unfortunately, heat-demanding ‘hot’ programming is generally used by almost every class of existing SMPs. By contrast, SMPs that can be ‘cold’ programmed (that is, deformed to a temporary shape at or below room temperature), which could greatly enhance the process ability to accommodate broader application requirements (for example, room temperature operations for the entire SM cycle), are rare. In addition, most of the current SMP applications focus on leveraging the macroscopic SM effects, where the deformation length scale is large (on the order of centimeters). However, an intriguing potential for all SMPs, largely unexplored, is their ability to memorize and change shape at nanoscale.3,12,14,29,72-75

Here, we report a new type of SMP that enables ‘cold’ programming and instantaneous, nanoscopic shape recovery at ambient conditions. No heat is needed for both SM programming and recovery steps. Instead the ‘cold’ programming is induced by

capillary pressure produced by water evaporation from the SMP membranes. Contrary to traditional thermoresponsive SMPs, the SM recovery of the new SMPs can be achieved by applying a small contact pressure or drying the films out of solvents with low surface tension (for example, ethanol). Interestingly, by combining this new type of

SMP with macroporous photonic crystals, we demonstrate the fabrication of reconfigurable/rewritable photonic crystal micropatterns by a simple direct print approach. Furthermore, in-situ nanoindentation tests reveal that the counterintuitive pressure-induced SM recovery is caused by an adhesive pull-off force between the contact substrate and the SMP membrane.

Methods

Templated Fabrication of Macroporous SMP Photonic Crystal Membranes.

Monodispersed silica microspheres, with diameter ranging from 100 to 600 nm, were

synthesized by the standard Sto¨ber method.76,77 Silica particles were purified in

200-proof ethanol by multiple centrifugation and redispersion cycles. Next, they were

self-assembled on a glass microslide by the convective self-assembly technology to

form colloidal crystals. By adjusting the particle volume fraction of the silica/ethanol

suspension, the thickness of the colloidal crystal was controlled to 10–50 colloidal

monolayers. The microslide with the silica colloidal crystal on its surface was covered

by another microslide and a double-sided adhesive tape of B1 mm thick was used as

a spacer in between the microslides. By using capillary force, the interstitial air in

between the silica microspheres was replaced by viscous oligomer mixtures

consisting of ETPTA (SR415, Sartomer, molecular weight 1,176 kDa, viscosity 225

cps at 25 C, refractive index 1.470) and PEGDA (SR610, Sartomer, molecular weight

742 kDa, viscosity 90 cps at 25 C, refractive index 1.468) oligomers with varying

volumetric ratios from 1:1 to 1:6. Darocur 1173 (2-hydroxy-2-methyl-1-phenyl-1-

propanone, BASF, 1 wt%) was added as the photoinitiator. The sample was

transferred to a pulsed ultraviolet curing system (RC 742, Xenon) and the oligomer

mixture was rapidly polymerized by exposure to ultraviolet radiation for 4 s. The

polymerized film was soaked in a 1 vol% hydrofluoric acid aqueous solution for 4 h

and then rinsed with deionized water. The resulting self-standing macroporous

polymer membrane showed pale iridescent colours when immersed in water and

observed at large viewing angles (≥ 45 ºC) due to the small refractive index contrast

between the copolymer (~ 1.47) and water (1.33).

Printing Photonic Crystal Patterns on Macroporous SMP Membranes.

Using Kim wipes, the free-standing macroporous SMP membrane was dried and the diffractive colours of the film were lost during water evaporation. Strikingly iridescent photonic crystal patterns, whose colours are determined by the size of the templating silica microspheres, can be printed on the translucent SMP membranes by using various substances with relief patterns, such as fingers or rubber stamps (see Fig. 2-3C,

E). To generate microscopic photonic crystal patterns, standard photolithography and chlorine reactive ion etching were performed in a class 100 cleanroom, to fabricate micropatterns (for example, parallel lines in Fig. 2-3G) on a silicon wafer. The hard silicon mould was then placed on an SMP membrane and a typical fingerprinting pressure was applied on the mould to transfer the micropatterns. To evaluate the pressure effects on the macroporous SMP strain-recovery rate, we placed different weights (43, 73, 85, 130, 162, 285 and 555 g) on a small polydimethylsiloxane square

(1 cm x 1 cm, Sylgard 184, cured at 75 C for 2 h), to generate various pressures on a macroporous copolymer membrane.

Sample Characterization.

SEM imaging was carried out on a FEI XL-40 FEGSEM. A thin layer of gold was sputtered onto the samples before imaging. Amplitude-modulation AFM was performed using a MFP-3D AFM (Asylum Research, Inc.) with a Nanosensor PPP-NCHR probe

(tip radius o10 nm), to characterize the topography and surface roughness of macroporous SMP membranes. In-situ nanoindentation tests were performed with an

MFP-3D NanoIndenter

(Asylum Research, Inc.) using a spherical sapphire indenter (tip radius ~125 mm). Such configuration of the instrument has a force and displacement resolution <3 mN and 1 nm, respectively. Detailed surface roughness analysis and calculation of

Young’s modulus by nanoindentation are discussed in Supplementary Information. An

Instron model 1122 load frame upgraded with an MTS ReNew system and equipped with a 500-g load cell at a cross-head speed of 0.5 mm /min was used in testing the tensile strength of the SMP membranes. Differential scanning calorimetric measurements were carried out from 80 to 18 ºC at a heating rate of 10 ºC /min using a

Seiko DSC 6200 instrument and an empty pan as reference. Normal-incidence optical reflection spectra were obtained using an Ocean Optics HR4000 high-resolution fiber optic visible-near infrared spectrometer with a reflection probe (R-400-7-SR) and a tungsten halogen light source (LS-1). Absolute reflectivity was obtained as the ratio of the sample spectrum and a reference spectrum, which was the optical density obtained from an aluminum-sputtered (1,000 nm thickness) silicon wafer. Scalar wave approximation optical modelling. The scalar wave theory developed for periodic dielectric structures was implemented to calculate the normal incidence optical reflection spectra from macroporous SMP photonic crystals. In this theory, Maxwell’s

equations are solved for a periodic dielectric assuming that one may neglect diffraction from all but one set of crystalline planes (for example, the (111) planes in our case). The scalar-wave approximation calculation contains no adjustable parameters, as the size of the macropores and the crystal thickness were independently determined from SEM characterization and the refractive index of the SMP copolymer is known.

Results

Preparation and Characterization of New SMPs.

The new pressure-responsive SMPs were discovered in the fabrication of macroporous polymer photonic crystal membranes. Photonic crystals are periodic dielectric structures with a forbidden photonic band gap (PBG) for electromagnetic waves.47,49,78-81 They may hold the key to continued progress towards all-optical integrated circuits and high-speed optical computing. Figure 2-1 compares the significant differences in SM effects between traditional thermoresponsive SMPs (Fig 2-

1A–C) and the new stimuli-responsive SMPs (Fig 2-1D–F). The new SMPs are photocured copolymers of ethoxylated (20) trimethylolpropane triacrylate (ETPTA, Tg~

40 ºC provided by the vendor) and polyethylene glycol (600) diacrylate (PEGDA, Tg~

42 ºC) oligomers (see molecular structures in Fig 2-2A) with varying volumetric ratios from 1:1 to 1:6. As the ETPTA-co-PEGDA copolymer with 1:3 ratio showed the optimal

SM behaviours, this recipe was adopted throughout the current work if not explicitly stated otherwise. A single Tg of ~ 42 ºC measured by differential scanning calorimetry

(Fig 2-2B) of an SMP sample indicates the cross-linked copolymer is a homogeneous mixture of the two components. The Young’s moduli of the pure ETPTA and PEGDA polymers and their 1:3 copolymers were characterized by in-situ nanoindentation tests

(see the typical force-depth indentation curve in Fig 2-2C). The results in Fig 2-2D show

that the average Young’s moduli for all samples are about 80 MPa. The tensile strength of the 1:3 copolymer membrane measured by using a conventional tensile tester is B7.5

MPa and the yielding strain is ~0.19 (Fig. 2-7), indicating the copolymer is quite elastic at room temperature. The bulk Young’s modulus is calculated to be B39 MPa, which is lower than the microscopic moduli measured by nanoindentation. Unusual ‘cold’ programming caused by water evaporation. Macroporous ETPTA-co-PEGDA membranes were fabricated by using self-assembled, three-dimensional (3D) highly ordered silica colloidal crystals as structural templates. After removing the templating silica microspheres by a hydrofluoric acid wash, the resulting macroporous copolymer film immersed in water exhibited iridescent colours caused by Bragg diffraction of visible light from the periodic arrays of polymer macropores. This confirmed the maintenance of the 3D ordered structure of the original silica colloidal crystal. Surprisingly, the shining colours of the macroporous photonic crystal disappeared when the membrane was dried out of water and it became translucent with a pale white appearance (Fig 2-3A).

This suggests that the 3D periodic structure was lost when water evaporated from the ordered macropores. The crosssectional scanning electron microscope (SEM) image in

Fig 2-3B confirms this conjecture, as no ordering of the deformed macropores is shown.

Therefore, the new elastic copolymers enable an autonomous ‘cold’ programming process—the deformation from a 3D ordered permanent structure to a disordered temporary structure can be achieved at ambient conditions by evaporating water from the templated macropores. This is in sharp contrast to traditional SMPs that need to be heated above T trans, then deformed to a temporary shape.

Pressure-Induced SM Recovery.

Even more interesting, the recovery of the permanent photonic crystal structure can be triggered at room temperature simply by applying a small contact pressure on the macroporous membranes with collapsed macropores. As illustrated by the fingerprinting process, an iridescent green-coloured fingerprint (Fig 2-3C) immediately appeared on the translucent macroporous copolymer membrane templated from 300 nm silica microspheres. The cross-sectional SEM image in Fig 2-3D shows a fingerprinted region with the vivid green colour and the recovered 3D highly ordered macroporous structure is evident. The difference in the surface microstructures between the fingerprint valleys and ridges was characterized by atomic force microscopy (AFM,

Fig 2-8). The surface of the raised fingerprint ridges is apparently much smoother than that of the valleys as confirmed by the magnified AFM images and the surface roughness analysis (Table 2-1). The raising height of the fingerprint ridges above the valleys is estimated to be ~1.5 mm by the AFM depth profile. The gradual transition from a disordered macroporous array in a fingerprint valley to a 3D highly ordered structure in a fingerprint ridge is shown by the SEM image in Fig 2-8,2-9. To avoid possible body-temperature effects on the macropore recovery in the above fingerprinting process, as well as to verify the feasibility of a new printing-based technology for fabricating arbitrary photonic crystal patterns, we printed a ‘light bulb’ relief pattern on a rubber stamp (Fig 2-3F) onto a translucent SMP copolymer membrane with collapsed 300 nm macropores at room temperature. The final iridescent imprint (Fig 2-3E) is a faithful replica of the original relief pattern. Furthermore, standard microfabrication technologies were used in making microscopic patterns on silicon wafers, which were used to imprint the micropatterns on SMP copolymer membranes.

Figure 2-3G shows an optical microscope image of printed pairs of parallel lines with

~30 mm width. The raising up of the smoother line patterns from the rough macroporous surface is clearly presented by the AFM image (Fig 2-3H) and the corresponding depth profile (Fig 2-3I).

Although various technologies for fabricating tunable photonic crystals have been demonstrated using elastic materials (for exmple, elastomers and gels), the temporarily deformed photonic microstructures cannot be memorized. Immediately, they return to the original crystalline lattices once the external stress is released. By contrast, the printed photonic crystal patterns on the new pressure-responsive SMPs are stable over long periods of time. The colourful fingerprints (for example, Fig 2-3C) stored at ambient conditions have maintained their vivid colours and clear patterns for more than 2 years.

Most importantly, the imprinted photonic crystal patterns can be erased when the SMP membranes are reimmersed in water and then dried out of it. New photonic microstructures can then be printed on the regenerated macroporous SMP membranes.

This unique rewriting capability is critical for developing reconfigurable photonic crystals that can adapt various photonic functionalities to accommodate different applications.

This reconfigurability can dramatically reduce the complexity and fabrication cost of developing a large number of application-specific devices. Capillary pressure-induced macropore collapse. Above we have shown that the new SMP copolymers enable room-temperature operations for the entire SM cycle (from an unusual ‘cold’ programming process to a contact pressure-induced recovery step). We speculated that the ‘cold’ programming process was induced by large capillary pressure created by water evaporation from the template macropores, which squeezed the elastic

macropores into disordered arrays. Similar macropore collapse was observed for macroporous polymer (for example, polysulfone) reverse osmosis membranes used for water purification. Further insight into the macropore collapse can be gained by considering the capillary pressure (Pc) in the Young–Laplace equation,

Pc = 2γcosθ/r, (2-1) where γ is the liquid/vapour surface tension, r is the radius of the pores and θ is the contact angle of the liquid on the pore surface. As the measured water contact angles on the copolymers were <20, cosθ is thus close to 1. One direct evidence supporting this capillary pressure-induced macropore collapse mechanism is that the templated macroporous membrane retained its original 3D ordered structure and iridescent colours when dried out of ethanol, which has a smaller surface tension than that of water (22.39 versus 72.75 mN m -1 at 20 ºC). The smaller γ led to a lower Pc that was not sufficient to squeeze the elastic macropores into disordered arrays. In addition to ethanol, a large variety of solvents with low surface tension (for example, acetone and toluene) can also trigger the same disorder-to-order transition. Figure 2-4A–D compare the surface microstructures of a macroporous SMP copolymer membrane dried out of water (Fig 2-4A, C) and ethanol (Fig 2-4B,D), respectively. The rough, disordered macroporous array was fully recovered to a smooth and ordered structure triggered by ethanol evaporation (Fig 2-4E). This disorder-to-order transition and the corresponding translucent-to-iridescent colour change can be characterized by measuring the normal- incidence optical reflection spectra. In Fig 2-4F, the sample with disordered macropores

(dried out of water) shows no apparent PBG peaks, whereas the ethanol recovered sample with ordered macropores exhibits a distinct PBG peak with well-defined Fabry–

Perot fringes, indicating the high crystalline quality of the solvent-activated macroporous photonic crystal. Importantly, the experimental spectrum of the recovered photonic crystal matches well with the calculated spectrum using a scalar-wave approximation model, which assumes a perfect macroporous crystalline lattice. This indicates that the temporarily deformed macropores were fully reopened to their permanent, 3D highly ordered structure triggered by ethanol evaporation. Critical contact pressure for SM recovery. By using the ethanol activated sample as a fully recovered control, we can estimate the critical pressure that is needed to trigger the shape recovery and the nanoscopic strain recovery rate (Rr) of the new SMPs through monitoring the PBG properties of the recovered samples under different pressures. Figure 2-5A shows that different reflection amplitudes of the PBG peaks resulted when various pressures were applied by putting varying weights on a small polydimethylsiloxane piece with a specific area on a macroporous copolymer membrane. Previous work has shown that the PBG optical density of a macroporous photonic crystal is nearly a linear function of its crystalline thickness. We therefore normalized the absolute reflection amplitude (after spectrum baseline correction) of a pressure-recovered SMP sample to that of the ethanol-activated control sample as an indicator of Rr (Fig 2-5B). A near-unity Rr was obtained when a 54.4-kPa pressure was applied, whereas a B50% recovery needed a pressure of 4.21 kPa. The cross-sectional SEM images in Fig. 5C,D, which correspond to the samples recovered by applying 7.13 and 27.9 kPa pressure, responsively, reveal that an intermediate pressure only induces the partial recovery of the top layers of the macroporous SMP photonic crystal, leading to the lower reflection amplitude compared with the fully recovered control sample.

Pressure-Induced SM Recovery Mechanisms.

Similar to thermoresponsive SMPs, we believe the entropy elasticity is the energetic root cause for the SM effects exhibited by the new pressure-responsive

SMPs. When photocured in the interstitial regions of the templating colloidal crystal, the cross-linked polymer chains were primarily in energetically favourable, stress-free configurations. The capillary pressure-induced ‘cold’ programming squeezed the ordered macropores into disordered arrays with reduced thickness (see Fig 2-3B,D), storing excess stresses in the deformed, temporarily configured polymer chains. The strained polymer networks have a strong tendency to recover back to their permanent, stress-free states. However, the observed pressure-induced macropore recovery is counterintuitive as we expect the applied pressure should further deform the collapsed macropores instead of popping them up. To elucidate this unusual shape-recovery mechanism, we conducted in-situ nanoindentation tests to characterize the forces in the approaching and retracting processes when a spherical sapphire tip indented the macroporous SMP membrane (Fig 2-6). An apparent adhesive pull-off force, caused by the attractive van der Waals interactions and the capillary force induced by the capillary- condensed water meniscus layer between the indenter tip and the SMP membrane, is evident in the retraction process. We believe this pull-off force causes the SM recovery of the collapsed macropores. A higher pressure leads to more conformal interactions between the molecules on the tip and the membrane, and thus a larger pull-off force.

This could explain the pressure effects on the strain recovery rate (Fig 2-5B) and the partial recovery of the top-layer macropores when an intermediate pressure was applied

(Fig 2-5C). One strong evidence supporting this pull-off mechanism is that the collapsed macropores did not recover back to their ordered structure when a pressure (up to~350

kPa) was applied through compressed air instead of a contacting body or entity (for example, a finger or a rubber stamp), which could exert the pull-off force on the SMP macropores.

Young’s Moduli of the New Pressure-Responsive SMPs Determined by Nanoindentation

Nanoindentation can be used to determine the Young’s modulus of a broad range of materials. By pressing an indenter into a material with a predefined depth or force, a force-displacement curve can be generated. By fitting the curve with an appropriate contact mechanics model (i.e. Oliver-Pharr model), material properties such as Young’s modulus and hardness can be extracted. Indentation tests were performed with a MFP-3D NanoIndenter (Asylum Research, Inc.) using a spherical sapphire indenter (tip radius ~125 m). Such configuration of the instrument has a force and displacement resolution less than 3 N and 1 nm, respectively. Due to the comparatively large contact radius of the spherical tip, there was no need to perform a tip area calibration according to Oliver and Pharr1. Tip geometry was directly measured by optical microscope. A force controlled trapezoidal load function with a 5-2-2 seconds segments corresponding to loading-hold-unloading times was applied to all indentations. Three forces (100 N, 200 N, and 300 N) were chosen to compare the

Young’s modulus of different indentation forces/depths. With each force, ten impressions were indented on each sample. Overall, 30 indents were made on each sample. All indents were made at room temperature (23 C) and the system was allowed to reach thermal equilibrium for 30 minutes prior to indentation to minimize the thermal drift effect.

To calculate the Young’s modulus, the force-displacement curves obtained from indentation experiments were fitted with the Oliver-Pharr model1 in 80%  20% portion of the unloading curves. The calculations were conducted using IGOR Pro analysis software routine (WaveMetrics Inc.). The fitting curves are in power law function form:

푚 푃 = 훼(ℎ − ℎ푓) (2-1) where α and m are power law fitting constants. hf is the final depth of the contact impression after unloading.

The contact depth is defined as the difference of the maximum indentation depth and the sink-in depth

푃 ℎ = ℎ − 휖 푚푎푥 (2-2) 푐 푚푎푥 푆 where ϵ is the indenter geometry parameter and S is the measured unloading stiffness, which is defined as

2 푆 = 훽 퐸푒푓푓√퐴푐 √휋 (2-3) where 훽 is a dimensionless parameter used to account for deviations in stiffness caused by lack of axial symmetry, Ac is the projected contact area of the indenter with repect to contact depth hc,

2 퐴푐 = −휋ℎ푐 + 2휋푅ℎ푐 (2-4) and Eeff is the effective (reduced) Young’s modulus defined by

1 1 − 휐2 1 − 휐2 (2-5) = 푠 + 푖 퐸푒푓푓 퐸푠 퐸푖 where Es, νs and Ei, vi are Young’s modulus and Poisson’s ratio of sample and indenter, respectively. R is the indenter tip radius.

According to the Oliver-Pharr method82 and the manufacturer‘s specifications, several parameters used in our case were set as 휖 = 0.75, 훽 = 1.05, 휈푠 = 0.5, 휈푖 =

0.29, 푎푛푑 퐸푖 = 350 퐺푃푎.

AFM Surface Microstructure Characterization and Roughness Analysis

Amplitude-modulation atomic force microscopy (Asylum Research, Inc.) was used to characterize the topography of the macroporous SMP membranes. All AM-AFM imaging was performed using the MFP-3D AFM with a Nanosensor PPP-NCHR probe

(tip radius < 10 nm). A total of 5 images were scanned in different locations on each sample with an average interval larger than 100 μm. All the 3 × 3 μm2, 10 × 10 μm2 and

90 × 90 μm2 images were scanned with a data collection density of 512 × 512 pixels per image. For all images presented, the trace and retrace images of the topography matched excellently, ensuring the absence of image artifacts and the accuracy of the data collected. Both the sample preparation and imaging were performed at room temperature (~ 23 °C) and relative humidity ~ 50%.

All the surface topographic images and the surface roughness were generated and calculated in the commercial software package Scanning Probe Imaging Processor

(SPIP). A 1st order plane correction was performed to compensate for surface tilt.

Cross-sectional profiles were measured from AFM imaging data to provide quantitative information, such as feature heights and lengths. The AFM images and the height profile scanned across the dashed line shown in Fig 2-7 illustrate the difference in the surface microstructures between the fingerprint valleys and ridges printed on a macroporous SMP membrane with 300 nm macropores.

The surface roughness was determined by both the arithmetic average (AA) roughness 푅푎 and the root mean square (RMS) roughness 푅푞 using AFM topographic images of sample. Each average roughness datum was calculated from 15 roughness values  3 samples with 5 different locations on each. Two different approaches were implemented to characterize the difference between the whole area (including pores) roughness and only non-porous area roughness:

3D areal roughness: This method includes whole surface area and porous features.

Linear profile roughness: This method extracts out several lines without porous features.

The roughness values are calculated according to the ASME B46.1. The formulas used in calculating the roughness are:

푛 1 (2-6) 푅 = ∑|푦 | 푎 푛 푖 푖=1

푛 1 푅 = √ ∑ 푦2 (2-7) 푞 푛 푖 푖=1

1 푆 = ∑푚 ∑푛 |푧(푥 , 푦 )| (2-8) 푎 푚푛 푘=1 푙=1 푘 푙

1 푆 = √ ∑푚 ∑푛 |푧(푥 , 푦 )|2 (2-9) 푞 푚푛 푘=1 푙=1 푘 푙

For our case, m = n = 512. Table 2-1 summarizes the surface roughness results obtained from macroporous SMP membranes dried out of water and ethanol, and the fingerprint valley and ridge regions.

Discussion

To conclude, we have developed a new type of stimuli-responsive SMP that differs greatly from existing SMPs, as it enables fast response and room-temperature operations for the entire SM cycle. The large capillary pressure generated by water evaporation from the templated macropores induces the unusual ‘cold’ programming at room temperature. The instantaneous recovery of the temporarily deformed macropores to the permanent, 3D highly ordered photonic crystal structure can be triggered by applying a small contact pressure at ambient conditions. In addition to being pressure responsive, the disorder-to-order transition of the new SMPs can also be stimulated by drying the macroporous SMP membranes out of solvents with low surface tension, such as ethanol and toluene. Importantly, the easily perceived colour change from translucency to iridescence associated with the structural disorder-to-order transition enables a simple and quantitative optical technology for characterizing the intriguing SM effects at nanoscale. Simultaneously, the striking chromogenic effects induced by the recovery of the permanent 3D photonic crystal structure provide opportunities for a wide spectrum of applications ranging from reconfigurable photonic crystal devices to chromogenic pressure and chemical sensors to novel biometric and anti-counterfeiting materials.

Figure 2-1. Schematic illustration compares the SM effects between a traditional thermoresponsive SMP and the new pressure-responsive SMP. A) A bulk SMP sample can be deformed into a temporary shape by a "hot" programming process (at a temperature above a specific Ttrans). B) The temperature is cooled below Ttrans to "freeze" in the temporary shape. C) The recovery of the permanent shape can be triggered by applying heat (T > Ttrans) to the strained SMP sample. D) A thin (a few m thick) macroporous SMP photonic crystal with 3-D ordered macropores (permanent shape) shows strong Bragg diffraction of visible light. E) The ordered macropores can be deformed to a disordered structure (temporary shape) with no Bragg diffraction induced by an autonomous "cold" programming process at ambient conditions. F) The nanoscopic recovery of the permanent 3-D photonic crystal structure can be stimulated by applying an external contact pressure.

Figure 2-2. Structural, thermal, and mechanical properties of the new pressure- responsive SMPs. A) Molecular structures of ETPTA 20 and PEGDA 600 oligomers. B) DSC plot of a macroporous ETPTA 20-co-PEGDA 600 copolymer with 1:3 ratios. C) Typical force-depth indentation curve used to measure elastic properties of SMP. D) Average Young's moduli for PEGDA 600, ETPTA 20-co-PEGDA 600 copolymer with 1:3 ratios, and ETPTA 20 polymers. Photo courtesy of author.

Figure 2-3. Arbitrary photonic crystal patterns printed on the new pressure-responsive SMP membranes. A) Photograph of a translucent macroporous SMP film with disordered macropores. B) Cross-sectional SEM image of the macroporous sample in A) with deformed macropores. C) Photograph of a green-colored fingerprint pressed on the sample in A). D) Cross-sectional SEM image of an iridescent region in C) with 3-D ordered macropores (300 nm diameter). E) Photograph of an iridescent "light bulb" pattern printed on a translucent macroporous SMP membrane templated from 300 nm silica microspheres. F) Photograph of the rubber stamp used in generating the "light bulb" pattern in E). G) Optical microscope image of micropatterned pairs of double lines on a macroporous SMP membrane. H, I) 3-D AFM image and the height profile of a section of a line in G). Photo courtesy of author.

Figure 2-4. Difference in topography and PBGs between a macroporous SMP membrane dried out of water and ethanol. A,C) 3-D AFM image and the height profile scanned across the profile line for a water-dried SMP sample consisting of 280 nm macropores. B,D) 3-D AFM image and the height profile for the same sample dried out of ethanol. E) Cross-sectional SEM image of the sample in B). F) Normal-incidence optical reflection spectra comparing the PBG properties of the macroporous samples dried out of water and ethanol. The SWA-simulated spectrum assuming a perfect macroporous crystalline lattice is also shown to compare with the experimental results. Photo courtesy of author.

Figure 2-5. Pressure-dependent macropore recovery. A) Normal-incidence optical reflection spectra obtained from a macroporous SMP membrane consisting of 280 nm macropores under different pressures. The same sample dried out of ethanol was labeled as "Full Recovery". B) Normalized absolute reflection amplitude of a recovered SMP sample was used as an indicator of the nanoscopic SMP strain recovery rate under different pressures. C,D) Cross- sectional SEM images of a macroporous SMP membrane recovered by 7.13 and 27.9 kPa pressure, respectively. Photo courtesy of author.

Figure 2-6. Macropore recovery induced by pull-off forces. A typical indentation force- displacement curve showing approach and retraction segments and pull-off force obtained on a macroporous SMP membrane. Inset show diagram of macropore recovery during retraction of indenter due to pull-off force caused by van der Waals interactions and the capillary force induced by the capillary- condensed water meniscus layer between the indenter tip and the SMP membrane.

Figure 2-7. Typical tensile stress versus strain curve for an ETPTA-co-PEGDA copolymer membrane with 1:3 ratios.

Figure 2-8. 3-D AFM images of a fingerprinted SMP sample with cross-sectional profiles. Inset: (Left) The 3 × 3 μm2 AFM image inside the valley region of the 90 × 90 μm2 fingerprint sample surface. (Right) The 3 × 3 μm2 AFM image inside the raised ridge region of the 90 × 90 μm2 fingerprint sample surface. (Middle) The cross-sectional profile of the 90 × 90 μm2 fingerprint sample surface.

Figure 2-9. Cross-sectional SEM image showing the transition from the disordered fingerprint valley region to the 3-D ordered fingerprint ridge region. Photo courtesy of author.

Table 2-1. Roughness of 10  10 m2 AFM scan size of SMP surface. Sample 3D Areal Roughness Linear Profile Roughness AA RMS RMS AA Roughness Roughness Sa Roughness Sq Roughness Rq Ra (nm) (nm) (nm) (nm) Water-dried 46.5 ± 7.1 59.4 ± 9.9 34.6 ± 6.3 41.7 ± 7.8 Ethanol-dried 8.7 ± 2.5 11.9 ± 4.3 6.3 ± 1.2 7.7 ± 1.3 FP valley 64.2 ± 21.6 84.9 ± 26.7 47.4 ± 15 56.5 ± 17.6 FP ridge 11.8 ± 2.9 16.4 ± 4.4 8.7 ± 2.1 10.9 ± 2.6

CHAPTER 3 CHROMOGENIC PHOTONIC CRYSTALS ENABLED BY NOVEL VAPOR- RESPONSIVE SHAPE MEMORY POLYMERS

Introduction

Shape memory polymers (SMPs) are a class of smart materials that can recover back to their "memorized" permanent shapes from temporary configurations in response to an external stimulus, such as heat, light, solvent, electric and magnetic fields.8,13,17,18,24,28,75,83-88Compared with their alloy counterparts (e.g., nitinol alloy), SMPs have gained increased attention due to their dramatically larger strain storage and recovery (up to 800% vs. less than 8%), light weight, low cost, ease of synthesis, and biocompatibility. They have been extensively explored for a wide spectrum of technological applications, such as reconfigurable morphing structures, smart textiles, sensors and actuators, self-healing materials, surgical stents and sutures, and implants for minimally invasive surgery. 6,28,30,39,71,89-96Traditional thermoresponsive shape memory (SM) effect is usually achieved in three steps including programming, storage, and recovery. Programming involves deforming a bulk SMP sample from its permanent shape to a temporary shape at a temperature higher than some specific transition temperatures (Ttrans) of the polymer, such as melting temperature (Tm) or glass transition temperature (Tg). The deformed sample is then cooled below Ttrans to fix the temporary shape which can be stored at ambient conditions for a long period of time. Recovery to the permanent shape, which is caused by entropy elasticity, occurs when the sample is reheated to above Ttrans.

The recovery time for bulk thermoresponsive SMPs, which are mostly studied and employed in practical devices, is usually long. This significantly impedes many applications that require fast response speed. Similar slow SM response is also suffered

by many other types of SMPs activated through laser, solvent, electric field, infrared absorption and alternating magnetic field. Indeed, most of these different SM activation mechanisms are still thermoresponsive as they depend on the generation of heat by various means to trigger the final shape recovery. Additionally, "hot" programming is generally utilized by almost every class of existing SMPs. By contrast, SMPs that can be "cold" programmed (i.e., deformed to a temporary shape at or below room temperature), which could greatly enhance the processability to accommodate broader application requirements (e.g., room temperature operations for the entire SM cycle), are rare. Moreover, most of the current SMP applications leverage the macroscopic SM effects and the deformation length scale is usually large (on the order of centimeter or larger). However, an intriguing potential for many new applications, largely unexplored, is the ability of SMPs to memorize and change shape at nanoscale. Furthermore, although a variety of solvents (e.g., water) can trigger SM recovery by effectively reducing Tg of the polymer through the plasticizing effect, vapor-responsive SMPs are uncommon.

Methods

Fabrication of Macroporous SMP Photonic Crystal Membranes

The synthesis of monodispersed silica microspheres with less than 5% diameter variation was performed by following the well-established Stöber method.76,77The synthesized silica microspheres were purified in 200-proof ethanol by multiple (at least 6 times) centrifugation and redispersion cycles. The purified silica particles were then assembled into 3-D highly ordered colloidal crystals on glass microslides using the convective self-assembly technology.76,97 The microslide with the silica colloidal crystal on its surface was covered by another microslide, separated by a double-sided

adhesive tape spacer (~ 1.7 mm thick). By utilizing capillary force, the interstitials inbetween the assembled silica microspheres were filled up with viscous oligomer mixtures consisting of ethoxylated (20) trimethylolpropane triacrylate (SR415, Sartomer) and polyethylene glycol (600) diacrylate (SR610, Sartomer) oligomers with varying volumetric ratios from 1:1 to 1:6. Darocur 1173 (2-hydroxy-2-methyl-1-phenyl-1- propanone, BASF, 1 wt %) was added as photoinitiator. The monomer mixture was photopolymerized by using a pulsed UV curing system (RC 742, Xenon) for 4 s. The solidified film was soaked in a 1 vol % hydrofluoric acid aqueous solution for 4 h and finally rinsed with deionized water.

Responses of Macroporous SMP Membranes Exposed to Acetone Vapors with Different Partial Pressures:

The templated macroporous SMP photonic crystal membrane was placed horizontally in a home-made environmental chamber. A reflection probe (R600-7,

Ocean Optics) connected to an Ocean Optics HR4000 high-resolution vis-NIR spectrometer was sealed in the environmental chamber to measure the optical reflectance from the SMP photonic crystal. The chamber was first purged with pure nitrogen gas for 2 min. It was then filled up with acetone vapors with different pressures.

Dry nitrogen was used to control the total pressure of the chamber to be 1 atm.

Sample Characterization:

SEM imaging was carried out on a FEI XL-40 FEG-SEM. A 15 nm thick gold layer was sputtered onto the samples prior to imaging. Amplitude-modulation atomic force microscopy (AM-AFM) was performed uing a MFP-3D AFM (Asylum Research,

Inc.) with a Nanosensor PPP-NCHR probe (tip radius < 10 nm). Differential scanning calorimetric measurements were performed from 80 to 18 C at a heating rate of 10 C

min-1 using a Seiko DSC 6200 instrument and an empty pan as reference. Normal- incidence optical reflection spectra were obtained using the Ocean Optics HR4000 high- resolution vis-NIR spectrometer with the R600-7 reflection probe and a tungsten halogen light source (LS-1). Absolute reflectivity was obtained as the ratio of the sample spectrum and a reference spectrum, which was the optical density obtained from an aluminum-sputtered (1000 nm thick) silicon wafer.

Nanoindentation Tests:

Nanoindentation tests were performed with a MFP-3D NanoIndenter (Asylum

Research, Inc.) using a spherical sapphire indenter (tip radius ~ 125 µm). Such configuration of the instrument has a force and displacement resolution less than 3 µN and 1 nm, respectively. Due to the comparatively large contact area of the spherical tip, there is no need to perform a contact area calibration. A force controlled trapezoidal load function with a 5-2-2 seconds segments corresponding to loading-hold-unloading times was applied to all indentations. Three forces, 100 µN, 200 µN, and 300 µN, were chosen to compare the Young’s modulus of different indentation forces/depths. With each force, ten impressions were indented on each sample with an inter-distance of 200

µm, which is about ten times over the average residual impression size. All indentations were triggered by 7.5 µN force, corresponding to ~ 2 nm deflection in the indenter spring. Overall, 30 indents were made on each sample. All indents were made at room temperature (23 C) and the system was allowed to reach thermal equilibrium for 30 minutes to minimize the thermal drift effect.

Scalar Wave Approximation Modeling:

The scalar wave theory developed for periodic dielectric structures, which solves

Maxwell’s equations by neglecting diffraction from all but one set of crystalline planes

(e.g., the (111) planes in this study), was utilized to calculate the normal-incidence optical reflection spectra from macroporous SMP photonic crystals. The structural parameters of the photonic crystals used in the optical modeling, including the size of the macropores and the crystalline thickness, were derived from SEM images. The refractive index of the SMP copolymers was assumed to be 1.47.

Results

Here we report a new type of vapor-responsive SMP that enables unusual "cold" programming and instantaneous shape recovery at nanoscale. These novel SMPs were discovered in the fabrication of macroporous polymer photonic crystals.97 Photonic crystals are periodic dielectric structures with a forbidden gap (or photonic band gap) for electromagnetic waves, analogous to the electronic band gap in semiconductors which lies at the heart of microelectronics. Photons with energies lying in the photonic band gap (PBG) cannot propagate through the medium, providing unprecedented opportunities to control the flow of light in miniature volumes for a large variety of applications ranging from all-optical integrated circuits to diffractive optical devices (e.g., optical filters). 98-104Tunable photonic crystals, whose lattice structures and PBGs can be adjusted by various stimuli, such as external pressure, electric and magnetic fields, solvents, vapors and metal ions, have been extensively investigated by using elastic materials (e.g., elastomers and gels).48,105-114 However, the deformed photonic crystal structures cannot be memorized and they rapidly return to the original crystalline lattices once the external stimuli are released. Although smart SMPs could provide a unique opportunity to realize reconfigurable photonic crystals with bistable states

(corresponding to the permanent and the temporary shapes of a SMP), these stimuli- responsive materials have rarely been used in previous photonic crystal studies.

By integrating scientific principles drawn from two disparate fields that do not typically intersect  the fast-growing photonic crystal and SMP technologies, here we demonstrate that reconfigurable photonic crystals exhibiting striking chromogenic effects can be achieved by using a new type of vapor-responsive SMP. Interestingly, our recent work has shown that the SM recovery of the same type of SMP can also be rapidly triggered by applying an external contact pressure. As illustrated by the scheme in Figure 3-1, the permanent photonic crystal configuration used in the current work is a three-dimensional (3-D) periodic array of macropores. These macroporous SMP photonic crystals were farbricated by using 3-D highly ordered silica colloidal crystals as structural templates. The templating colloidal crystals were assembled by the convective self-assembly technology using silica microspheres with diameter ranging from 200 to 400 nm. The thickness of the resulting colloidal crystal was controlled to ~

35 m (or ~ 1020 colloidal monolayers) by adjusting the concentration of the colloidal suspensions in the convective self-assembly process. The interstitial air in-between the silica microspheres was replaced by viscous oligomer mixtures of ethoxylated (20) trimethylolpropane triacrylate (ETPTA 20, MW 1176, viscosity 225 cps at 25 C, Tg ~

40 C, refractive index ~ 1.470) and polyethylene glycol (600) diacrylate (PEGDA 600,

MW 742, viscosity 90 cps at 25 C, Tg ~ 42 C, refractive index ~ 1.468) with varying volumetric ratios from 1:1 to 1:6. The molecular structures of these oligomers are shown in Figure 3-7. The oligomer mixture was then photocured at ambient conditions and the templating silica microspheres were selectively dissolved in a hydrofluoric acid aqueous solution, leaving behind a free-standing macroporous ETPTA 20-co-PEGDA 600 copolymer membrane with crystalline arrays of macropores. The size of the templated

macropores, which determines the color of the final macroporous photonic crystal, is defined by the diameter of the templating silica microspheres. Differential scanning calorimetric (DSC) measurements (see a typical DSC plot of a macroporous ETPTA 20- co-PEGDA 600 copolymer film with 1:3 ratios in Figure 3-8) show the copolymers have

Tg close to those of the two oligomer components (~ 42 C), indicating the crosslinked copolymers are rubbery at room temperature. When immersed in water, the templated macroporous SMP membrane shows pale iridescent colors at large viewing angles (>

45 ) caused by Bragg diffraction of visible light from the periodic arrays of polymer macropores filled with water (refractive index 1.333). This confirms the maintenance of the 3-D ordered structure of the original silica colloidal crystal throughout the templating process.

The unusual "cold" programming process occurred when the macroporous SMP membrane dried out of water. Surprisingly, the iridescent color of the macroporous photonic crystal disappeared, and the film became translucent with a pale white appearance (Figure 3-2A). The typical cross-sectional scanning electron microscope

(SEM) image of the water-dried sample in Figure 3-2B shows no apparent ordering of the templated macropores, indicating an order-disorder transition during water evaporation. The atomic force microscopy (AFM) image (Figure 3-3C) and the depth profile scanned across the line (Figure 3-2D) illustrate that the surface of the dried membrane is rough. The root-mean-square (RMS) linear profile roughness (Rq) of the sample was determined to be 41.7  7.8 nm (Table 3-1). We attributed this order- disorder transition during water evaporation to the large capillary pressure induced by the high surface tension of water (72.75 mN/m at 20 C), which is sufficient to compress

the ordered elastic macropores into disordered arrays. According to the Young-Laplace equation,

Pc = 2 cos / r, (3-1) the capillary pressure (Pc) is proportional to the liquid/vapor surface tension () and cos

( is the contact angle of the liquid on the pore surface); while inversely proportional to the radius of the pores (r). Therefore, Pc can be reduced by using a solvent with a low surface tension (e.g., ethanol with  ~ 22.39 mN/m at 20 C), or by increasing the size of the pores. When Pc is small compared with the elastic modulus (or Young's modulus, E) of the copolymer membrane, we expect the templated SMP macropores will remain the original 3-D highly ordered structure, instead of being squeezed into disordered arrays.

This speculation was supported by two experimental evidences. First, iridescent macroporous copolymer membranes with striking diffractive colors and much smoother surface (Rq ~ 7.7  1.3 nm) resulted when the macroporous SMP samples dried out of ethanol instead of water. Second, our experiments showed that SMP membranes templated from large silica microspheres (> 600 nm) maintained the ordered structure even when dried out of water.

We evaluated the Young's modulus of the macroporous SMP membranes by in- situ nanoindentation tests. Three forces (100 N, 200 N, and 300 N) were chosen to compare E of different indentation forces/depths. Figure3-9 shows the results of macroporous ETPTA 20-co-PEGDA 600 (1:3 ratio) membranes with 300 nm macropores dried out of ethanol and water, respectively. Apparently, all samples have

Young's modulus of ~ 30 MPa, and the water-dried films are slightly stiffer than the ethanol-dried ones. This is reasonable as more air was trapped in the ethanol-dried

samples with 3-D ordered macropores compared with the collapsed pores of the water- dried membranes. Similar E values were obtained from macroporous copolymer samples with other oligomer volumetric ratios (from 1:1 to 1:6) and macropore sizes. By using the Young-Laplace equation, we estimated the capillary pressure generated by evaporating water from the copolymer macropores with 300 nm diameter to be ~ 1 MPa

(i.e., ~ 10 atm), which is comparable with the Young's modulus of the macroporous

SMP membranes. Similar capillary pressure-induced macropore collapse has been reported during drying of macroporous polymer (e.g., polysulfone) reverse osmosis membranes used for water purification.

The autonomous evaporation-assisted "cold" programming exhibited by the macroporous SMP membranes is in sharp contrast to the common "hot" programming process used by traditional SMPs. Even more interesting, a translucent macroporous

SMP copolymer membrane with collapsed macropores momentarily changed color from pale white to brilliant iridescence (Figure 3-3A) when the sample was exposed to various organic vapors (e.g., acetone, methanol, and chloroform) at ambient conditions.

The cross-sectional SEM image in Figure 3-3B shows a SMP copolymer sample after exposing to an acetone vapor. The recovery of the 3-D highly ordered photonic crystal structure (permanent configuration) is evident. By averaging over 50 different spots on a few SEM images, the thickness of the macroporous layers of the water-dried and the acetone vapor-recovered SMP samples was estimated to be 1.95  0.13 m and 5.75 

0.06 m, respectively. The nearly 3-fold expansion of the macroporous layer indicates the collapsed macropores popped up into ordered arrays when triggered by acetone vapor exposure. The AFM image and the depth profile in Figure 3-3C-D illustrate that

the acetone-recovered macroporous SMP membrane has a much smoother surface than the water-dried sample (Figure 3-2C-D). Table 1 compares the surface roughness of a SMP membrane dried out water, ethanol, and acetone vapor. The acetone vapor- activated sample has a slightly rougher surface than the liquid ethanol-recovered one; while both samples are significantly smoother than the water-dried membrane.

The instantaneous transition between a disordered temporary configuration and a 3-D highly ordered permanent structure, which leads to an easily perceived color change from translucence to striking iridescence, can be quantitatively characterized by measuring the normal-incidence reflection spectra using an optical spectrometer. Figure

3-4 compares the optical reflection spectra obtained from a water-dried SMP membrane with 280 nm macropores (black line), and the same sample after exposed to acetone vapor (red line) and liquid ethanol (blue line). No apparent Bragg diffraction peaks are shown in the spectrum of the water-dried sample; while distinct diffraction peaks with well-defined Fabry-Perot fringes are present in the spectra of the samples triggered by acetone vapor and liquid ethanol, confirming the high crystalline quality of the recovered macroporus photonic crystals. Additionally, the experimental spectrum of the ethanol- recovered sample matches well with the calculated spectrum using a scalar-wave approximation (SWA) model, which assumes a perfect crystalline lattice. We can then use the ethanol-activated SMP membrane as a fully recovered control to evaluate the completeness of macropore recovery under different triggering conditions. As shown in

Figure 3-4, the amplitude of the PBG peaks of the acetone vapor-activated sample is slightly lower than that of the liquid ethanol-recovered one. As the PBG optical density of a macroporous photonic crystal is a sensitive function of its crystalline thickness, the

smaller reflection amplitude indicates the acetone vapor-triggered macropore recovery is not as complete as the liquid ethanol-induced recovery. This agrees with the surface roughness results shown in Table 3-1.

Above we have shown that the chromogenic responses enabled by the macroporous photonic crystals with micrometer-scale thickness provide a simple yet sensitive optical methodology for characterizing microscopic SM effects. We then used this optical tool to evaluate the reversibility, durability, and reproducibility of the vapor- triggered SMP membranes. Figure 3-5A and 3-5B compare the optical reflection spectra obtained from the same macroporous ETPTA 20-co-PEGDA 600 (1:3 ratios) membrane cyclically exposed to acetone vapor and then dried out of water for 10 times. The good reversibility and reproducibility of the sample are evident from the spectra and the comparison of the absolute reflection amplitude at 500 nm wavelength for the sample cyclically exposed to acetone vapor and water in Figure 3-5C. Indeed, our extensive tests showed that the macroporous SMP copolymer membranes could be reused for over 500 times without any apparent degradation in the chromogenic response to acetone vapor.

We speculated that the capillary condensation and evaporation of fluids with low surface tension in macroporous SMP membranes played a critical role in the vapor- triggered SM recovery. As shown by Figure 3-6A-C, the translucent SMP membrane instantaneously changed color to reddish when the sample was close to the surface of liquid acetone, where the partial pressure of acetone vapor was high. Interestingly, the sample could become nearly transparent when it stayed close to the liquid acetone surface for a while. This indicates all macropores were filled up with condensed acetone

whose refractive index (nacetone ~ 1.359) is close to that of the ETPTA 20-co-PEGDA 600 copolymer (ncopolym ~ 1.470). The reddish color changed to greenish when the membrane moved away from the liquid acetone surface, and this red-green color transition was reversible. To gain quantitative insights into the capillary condensation of condensable vapors in the macroporous SMP membranes, we measured the normal- incidence optical reflection spectra for a sample with 300 nm macropores exposed to acetone vapors with different partial pressures (Figure 3-6D). The diffraction peak red- shifts with increasing vapor pressure, and it nearly disappears (due to refractive index matching) when the vapor partial pressure is very high (682.8 mmHg). We can calculate

the effective refractive index ( neff ) of the macroporous photonic crystals with condensed liquid using the Bragg diffraction equation:

max = 2× neff × d × sin (3-2) where d is the inter-plane distance and  is /2 for normal incidence. By assuming the macropores are close-packed and the volume fraction of air (VFair) in a dry macroporous

SMP membrane is 0.74, we can then calculate the volume fraction of the condensed acetone (VFacetone) using neff = ncopolym  0.26 + nair  (0.74  VFacetone) + nacetone 

VFacetone, where ncopolym, nair, and nacetone is 1.47, 1.0, and 1.359, respectively. As shown in previous work, the condensed liquid forms a uniform thin layer on the walls of the maropores. The thickness of this liquid layer and the size of the remaining air cavities can be easily evaluated by using VFacetone. The calculated radius of air cavities for the 5 samples with apparent diffraction peaks and increasing vapor pressures in Figure 6d is

111.3, 99.8, 91.3, 81.9, sand 69.7 nm, respectively. We finally compared our experimental results with the predictions using the Kelvin equation,

P 2V ln  l (3-3) P0 rRT where P and P0 are actual and saturation vapor pressure,  is the liquid/vapor surface tension, Vl is the liquid molar volume, r is the radius of curvature. The Kelvin equation has been widely utilized in describing the phenomenon of capillary condensation due to the presence of a curved meniscus. It predicts lnP is linearly proportional to 1/r when other variables are constant. Our experimental results match well with this prediction

(inset of Figure 3-6D).

The overall shape memory cycle enabled by the novel vapor-responsive SMPs can be summarized as follows. When photopolymerized in the presence of the silica colloidal crystal template, the cross-linked polymer chains are in stress-free configurations which are energetically favorable. The large capillary pressure induced by the evaporation of water trapped in the templated macropores squeezes the 3-D ordered macropores into temporary disordered arrays. Excess stresses are stored in the deformed polymer chains and they tend to recover back to the original stress-free state due to entropy elasticity. The rapid capillary condensation of acetone vapors in the macropores triggers the instantaneous recovery of the permanent photonic crystal structure. As the surface tensions of the condensed liquids (e.g., acetone and methanol) are significantly lower than that of water, the evaporation-induced capillary pressure is not sufficient to deform the recovered macropores during capillary evaporation of the condensed liquids.

Conclusion

In conclusion, we have discovered a new type of vapor-responsive SMP that enables room-temperature operations for the entire shape memory cycle. The recovery

of the permanent macroporous photonic crystal structure can be momentarily triggered by a variety of organic vapors. The striking chromogenic effect (from colorless to iridescent) induced by the disorder-order transition differs greatly from the typical color change with limited wavelength shift exhibited by traditional tunable photonic crystals. In addition, the thin photonic crystal structure provides a simple yet sensitive optical technique for investigating the intriguing SM effects at nanoscale. These smart stimuli- responsive materials could find important technological applications ranging from reconfigurable nanooptical devices to reusable chromogenic vapor sensors.

Figure 3-1. Schematic illustration showing the SM effects of the new vapor-responsive SMP. A) Thin macroporous SMP photonic crystal with 3-D ordered structure can diffract light with specific wavelengths. B) The unusual "cold" programming process deforms the ordered macropores into disordered array with rough surface and no light diffraction. C) The recovery of the permanent photonic crystal structure can be triggered by exposing the deformed membrane to various organic vapors (e.g., acetone).

Figure 3-2. A macroporous SMP membrane with 280 nm macropores after drying out of water. A) Photograph of a macroporous SMP membrane with 280 nm macropores after drying out of water. B) Cross-sectional SEM image of the sample. C) AFM scan of the sample surface. D) Height profile of the dashed line in C). Photo courtesy of author.

Figure 3-3. A macroporous SMP membrane with 280 nm macropores after exposing to an acetone vapor. A) Photograph of a macroporous SMP membrane with 280 nm macropores after exposing to an acetone vapor. B) Cross-sectional SEM image of the sample. C) AFM scan of the sample surface. D) Height profile of the dashed line in C). Photo courtesy of author.

100 Collapsed macropores Acetone-recovered macropores Ethanol-recovered macropores 80 SWA simulation

Reflection (%) Reflection 20

0 400 500 600 700 800

Wavelength (nm)

Figure 3-4. Normal-incidence optical reflection spectra comparing a macroporous SMP membrane with 280 nm macropores dried out of water, liquid ethanol, and acetone vapor. The calculated spectrum using a SWA model is also shown to compare with the experimental results.

Figure 3-5. Normal-incidence optical reflection spectra obtained from a macroporous. A) Normal-incidence optical reflection spectra obtained from a macroporous SMP membrane with 280 nm macropores exposed to acetone vapor for 10 times. B) Normal-incidence reflection spectra of the same sample after drying out of water for 10 times. C) Reflection amplitudes of the spectra in A) and B) taken at 500 nm wavelength.

Figure 3-6. A macroporous SMP membrane exposed to acetone vapor above liquid acetone at different locations. A-C) Photographs showing a macroporous SMP membrane exposed to acetone vapor above liquid acetone at different locations. D) Normal-incidence optical reflection spectra obtained from a macroporous SMP membrane exposed to acetone vapors with different partial pressures. Inset showing dependence of lnP vs the reciprocal of the radius of curvature of the condensed acetone films. The pressure is in unit of mmHg and the radius of curvature is in unit of nm. Photo courtesy of author.

Figure 3-7. Molecular structure. A) Molecular structure of ETPTA 20 (x + y + z = 20). B) Molecular structure of PEGDA 600 (x = 12).

-7

-8

) -9

( -10 -42°C

-11

Heat Flow Heat Flow -12

-13 -80 -60 -40 -20 0 20

o Temperature ( C)

Figure 3-8. Typical DSC plot of a macroporous ETPTA 20-co-PEGDA 600 copolymer with 1:3 ratios.

40 Dried out of ethanol Dried out of water

1.76 2.73 1.75 1.15 30 2.19 1.36

Young's Modulus (MPa) Modulus Young's 0

100 N 200 N 300 N

Figure 3-9. Comparison of the Young’s modulus of the ethanol-dried and water-dried macroporous ETPTA 20-co-PEGDA 600 (1:3 ratio) membranes indented with different forces.

Table 3-1. Roughness of 10  10 m2 AFM scan area of SMP sample surface. 3-D Areal Roughness Linear Profile Roughness AA RMS AA RMS Samples Roughness Roughness Roughness Roughness Sa [nm] Sq [nm] Ra [nm] Rq [nm] Water-dried 46.5  7.1 59.4  9.9 34.6  6.3 41.7  7.8 Liquid ethanol- 8.7  2.5 11.9  4.3 6.3  1.2 7.7  1.3 activated Acetone vapor- 10.2  1.6 14.0  2.6 7.2  1.1 8.9  1.2 activated

CHAPTER 4 DIRECT WRITING OF THREE-DIMENSIONAL MACROPOROUS PHOTONIC CRYSTALS ON PRESSURE-RESPONSIVE SHAPE MEMORY POLYMERS

Introduction

Three-dimensional (3D) printing (or additive manufacturing) has attracted great recent interest as it enables rapid manufacturing and prototyping of 3D objects with arbitrary shapes and/or geometries. In 3D printing, successive layers of materials (e.g., polymers, ceramics, and metal alloys) are laid down under computer control through processes like inkjet printing, extrusion, and sintering. Beyond conventional manufacture of macroscopic objects (e.g., customized shoes, automobile parts, and even guns), 3D printing has also been extensively exploited for fabricating microscopic devices with unique optical, electrical, and magnetic properties.115-127 One preeminent example is the direct writing of 3D ordered photonic crystals with desired crystal structures and pre-engineered defects. Photonic crystals are periodic dielectric structures with a forbidden photonic band gap (PBG) for electromagnetic waves. As 3D photonic crystals with full PBGs can manipulate photons in a similar fashion as semiconductors do electrons, they provide enormous opportunities in controlling the flow of light in microscopic volumes for a plethora of applications ranging from all-optical integrated circuits and quantum information processing to low-threshold lasers and lossless waveguides. 101,128,129To fabricate photonic crystals possessing optical and near-infrared (NIR) PBGs, the lattice constant of the artificial crystal must have dimensions on the submicrometer scale.130 Unfortunately, this length scale is formidably challenging for direct-writing-based 3D printing technologies, especially considering the overflow of the ink materials (e.g., photopolymers) in the layer-by-layer deposition process.

Methods

Templating Nanofabrication of Macroporous SMP Photonic Crystal Membranes:

Monodispersed silica microspheres with diameter ranging from 200 to 400 nm were synthesized using the standard Stöber method.131 The as-synthesized silica microspheres were purified in 200-proof ethanol by using multiple centrifugation and redispersion cycles (at least 5 times). The purified silica particles were then self- assembled on glass microslides to form hexagonally close-packed colloidal single crystals using the well-established convective self-assembly technology. The thickness of the resulting colloidal crystal was controlled to ~ 5 µm by adjusting the particle volume fraction of the silica microspheres-ethanol suspension to ~ 1.0%. A double- sided adhesive tape (~ 1 mm thick) was used as a spacer to separate the glass microslide with the self-assembled silica colloidal crystal on its surface from another bare glass microslide. A viscous oligomer mixture containing 1.5 g polyethylene glycol

(600) diacrylate (SR610, Sartomer), 0.5 g ethoxylated (20) trimethylolpropane triacrylate

(SR415, Sartomer), and 0.016 g Darocur 1173 photoinitiator (2-hydroxy-2-methyl-1- phenyl-1-propanone, BASF) was injected in between the two glass microslides to completely fill up the gap. The sample became nearly transparent due to the refractive index matching between the oligomer mixture and the silica microspheres. The oligomers were then photopolymerized by exposing the sample to ultraviolet radiation for 4 s using a pulsed UV curing system (RC 742, Xenon). The solidified sample was finally soaked in a 1 vol % hydrofluoric acid aqueous solution for 4 h and then rinsed with deionized water. After blow-drying with compressed air, free-standing macroporous

SMP membranes were resulted.

Directly Printing and Hand-Writing 3D Macroscopic Photonic Crystal Patterns on Templated Macroporous SMP Membranes:

We prepared 2 × 2 cm2 macroporous SMP membranes as “writing pads” for directly printing and hand-writing 3D photonic crystal patterns on them. Commercial rubber stamps with relief patterns purchased from Office Depot were pressed gently for

2 s on the SMP “writing pads” to print colorful inverted patterns on the macroporous membranes. To directly write 3D photonic crystal features, a home-made writing tool was made by the normal pencil wrapped by Handi-wrap. Next, we can direct write 5mm x 5mm striking “UF” pattern with vivid greenish color on the ‘WordPad’ in 2 seconds. All the colorful pattern can be erased by deionized water and shows transparent again.

Through considerable investigation, the “WordPad” can be reused for more than 100 times without destructive.

Directly Writing 3D Photonic Crystal Micropatterns by AFM:

A MFP-3D atomic force microscope (Asylum Research, CA) was used for writing microscopic patterns on SMP membranes. Both the dedicated MFP-3D NanoIndenter

(flexure, k = 3814 N/m) module and the AFM cantilever-based configuration were used with a 1 mm (sapphire, E = 350 GPa) and a 20 µm (borosilicate, E = 62.8 GPa, nominal k ~ 42 N/m, length = 125 m, CP-NCH-BSG cantilever from sQUBE Inc., Bickenbach,

Germany) diameter spherical tip, respectively. The minimum force and the displacement resolution of the NanoIndenter module is less than 3 µN and 1 nm, respectively. The resolutions of the cantilever-based configuration are less than 6 nN and 0.1 nm. The writing forces of both configurations were controlled by closed-loop control of the set- point voltage, which defines the amount of the contact force maintained during writing.

The MicroAngeloTM software routine (Asylum Research) was used to program writing

parameters including setpoint, speed, feature geometry, etc. Detailed writing parameters and procedures using AFM are discussed in the Supporting Information.

Sample Characterization:

SEM imaging was carried out on a FEI Nova NanoSEM 430. A thin layer of gold was sputtered onto the samples prior to imaging. Amplitude-modulation atomic force microscopy (AM-AFM) was performed uing a MFP-3D AFM (Asylum Research, Inc.) with a Nanosensor PPP-NCHR probe (tip radius < 10 nm). All AFM images were processed using the Scanning Probe Imaging Processor (SPIP, Image Metrology Inc.,

Horsholm, Denmark) software. Normal-incidence optical reflection spectra were obtained using an Ocean Optics HR4000 high-resolution fiber optic vis-NIR spectrometer with a reflection probe (R600-7) and a tungsten halogen light source (LS-

1). Absolute reflectivity was obtained as the ratio of the sample spectrum and a reference spectrum, which was the optical density obtained from an aluminum-sputtered

(1000 nm thickness) silicon wafer.

Scalar Wave Approximation Optical Modeling:

The scalar wave theory developed for periodic dielectric structures was implemented to model the normal-incidence optical reflection spectra from macroporous

SMP photonic crystal membranes. In the SWA theory, Maxwell’s equations are solved for a periodic dielectric medium assuming that one may neglect diffraction from all but one set of crystalline planes. In the current work, only the (111) crystalline planes of a face-centered cubic crystal was considered in the modeling. The SWA simulation contains no adjustable parameters, as the size of the macropores and the crystal thickness were experimentally determined from SEM images, and the refractive index of the SMP copolymer was known.

Micropatterning with 1 mm Diameter Spherical Tip:

First, a series of lines were written with decreasing forces from 1400 to 3 N to determine the minimum contact force that could trigger the chromogenic SM recovery

(observed as shining green-colored reflective micropatterns). The spacing between the lines was 100 µm. The line widths were averaged by using optical microscopy and AFM images. The projected tip contact area (circular) was calculated using the measured line width as the circle diameter. The contact stress was calculated from the contact force divided by the contact area. The writing speed was held constant at 5 µm/s and the length of the lines was 90 µm. A “UF” pattern was also written to demonstrate the achievable writing resolution of letters with high quality. Both letters were smaller than

100 × 100 µm2. The “UF” letters were written using a 140 µN contact force and 5 µm/s writing speed. Zoomed-in scan of areas both inside and outside the “U” pattern was performed. The image size was 5 × 5 µm2 with a pixel resolution of ~ 20 nm.

Nanopatterning with 20 µm Diameter Spherical Tip:

A series of lines were written to study the writing speed dependence of the SM recovery with a 20 µm diameter spherical tip. The contact force throughout this set of experiments was 6 µN. It was chosen as the onset force that caused shining reflective patterns under the writing speed of 0.2 µm/s. The writing speed was increased from 0.2 to 20 µm/s with 3 lines written per speed. The spacing between the lines was 30 µm and the designed length of the lines was 90 µm.

Results

We report a single-step direct writing technology for reversibly printing 3D macroporous photonic crystal patterns (both macroscopic and microscopic) with submicrometer-scale lattice spacing on a new type of pressure-responsive shape

memory polymer (SMP). This technology integrates scientific principles drawn from two disparate fields  the well-established templating nanofabrication and shape memory materials.85 Compared with sequential 3D printing, which needs to resolve the resolution issue in generating the intrinsic 3D submicrometer-scale microstructures, the current approach utilizes colloidal crystal-based templating nanofabrication in defining the final photonic crystal lattice parameters. Self-assembled colloidal crystals have been widely used as structural template in fabricating macroporous photonic crystals with periodic arrays of air cavities embedded in the matrix material (e.g., polymer, metal, and semiconductor).97,103,132,133The stringent submicrometer-scale lattice spacing requirement for making visible- and NIR-active 3D photonic crystals can be easily satisfied by controlling the size of the templating colloidal particles.134 Another major merit of the current technology is the employment of new pressure-sensitive SMPs that enable the direct writing of arbitrary 3D macroporous photonic crystal patterns on the polymer surface in a single step. Shape memory polymers are a class of smart materials that can recover their “memorized” permanent shapes triggered by various external stimuli, such as heat, light, solvent, and electromagnetic field. Shape memory

(SM) effects in traditional SMPs are usually achieved in three steps  programming, storage, and recovery.135,136 In programming, a SMP sample is mechanically deformed from its permanent shape to a temporary configuration by heating the sample above a specific transition temperature (Ttrans), such as the polymer glass transition temperature (Tg). The temporary shape is then “frozen” in the polymer by cooling the deformed sample below Ttrans. Recovery to the permanent shape, which is caused by entropy elasticity, can finally be triggered by applying different stimuli, such as reheating

the sample above Ttrans or exposing it to ultraviolet radiation.137,138 Although thermoresponsive SMPs have previously been utilized in making tunable 3D colloidal photonic crystals and 2D diffractive gratings, the heat-demanding SM programming and recovery steps impede the ultimate performance and applications of the SMP-enabled microoptical devices.139,140

We have recently discovered a new type of stimuli-responsive shape memory polymer that enables unusual “cold” programming (i.e., the deformation from the permanent shape to the temporary configuration occurs at room temperature) and instantaneous shape recovery at ambient conditions triggered by applying an external contact pressure or exposing the polymer to various organic vapors (e.g., acetone and toluene). These new SMPs are composed of photocured copolymers of ethoxylated (20) trimethylolpropane triacrylate (ETPTA 20, Tg ~ 40 C, MW 1176, refractive index

1.470) and polyethylene glycol (600) diacrylate (PEGDA 600, Tg ~ 42 C, MW 742, refractive index 1.468) oligomers with varying volumetric ratios from 1:1 to 1:6. Figure 7 shows an exemplary pressure-induced SM recovery process using an ETPTA 20-co-

PEGDA 600 copolymer with a 1:3 volumetric ratios. The relief “+A” pattern on the surface of a commercial rubber stamp (Figure 4-7A) was inversely printed on a translucent SMP membrane with collapsed macropores (temporary configuration). The

SM recovery of the permanent, 3D ordered macroporous arrays, which were templated from self-assembled colloidal crystals consisting of 280 nm silica microspheres,97 led to the iridescent structural colors of the printed “A+” pattern in Figure 4-7B. To explain this counterintuitive pressure-induced recovery of collapsed macropores, we proposed a SM recovery mechanism triggered by an adhesive pull-off force caused by the attractive van

der Waals interactions the capillary force generated by the water meniscus bridging in between the rubber stamp and the pressure-responsive SMP membrane. However, this new recovery mechanism is far from being thoroughly investigated and verified. In this communication, we report a novel direct writing technology for inscribing arbitrary 3D photonic crystal patterns on the aforementioned pressure-responsive SMP membranes.

In sharp contrast to the vertical pull-off force in our previous static printing process (see

Figure 4-7), the lateral shear stress plays a critical role in this dynamic approach.

Quantitative insights into the unusual macropore recovery induced by the lateral interactions are gained through fundamental investigations on important process parameters, including the critical pressure and writing speed for triggering the recovery of the deformed macropores, and the minimal feature size that can be directly written on the SMP membranes.

The schemes in Figure 4-1 illustrate the basic concept of the new direct writing technology for making 3D ordered photonic crystal patterns on a macroporous SMP membrane with collapsed macropores. The self-standing SMP membranes were produced by templating nanofabrication using convectively self-assembled silica colloidal crystals as structural template. In this process, ETPTA 20 and PEGDA 600 oligomer mixtures were first photopolymerized in the interstitials of 3D ordered silica particle arrays. The cross-linked polymer chains in the 3D inversely ordered polymer matrix were primarily in an energetically favorable, stress-free configuration, denoting the permanent state of the SMPs. After removing the templating silica microspheres in a hydrofluoric acid aqueous solution and drying the SMP membrane out of water, the original ordered macropores were surprisingly collapsed, resulting in the translucent

appearance of the film (see Figure 4-7B). Our previous studies showed that strong capillary pressure induced by water evaporation deformed the elastic macropores (Tg of the copolymers << room temperature) into disordered arrays in this “cold” programming process, storing excess stresses in the squeezed, temporarily configured polymer chains. The recovery of the permanent, stress-free state (i.e., 3D ordered macroporous arrays) can be triggered by direct writing using either macroscopic or microscopic writing tools like a pen or an atomic force microscope (AFM) tip. As the SM recovery is confined only to the regions underneath the writing tool and the recovered feature size is mainly determined by the sharpness of the writing tip, we can generate nanoscopic photonic crystal patterns, like the letters “U” and “F” in Figure 4-1 (representing the abbreviation for University of Florida), using a sharp AFM tip. In addition to induce the above disorder-to-order transition, the direct writing process can also pop up the deformed macropores underneath the tip, making the recovered photonic crystal patterns protruding out of an otherwise disordered background.

We started to demonstrate the direct writing of 3D photonic crystal patterns on pressure-responsive SMP membranes using macroscopic writing tools like a conventional fountain pen (without ink). However, the direct writing-induced SM recovery of collapsed macropores was not as straightforward as that exhibited by static printing. Although well-defined writing marks were left underneath the stainless steel tip, these marks were pale-colored, indicating an incomplete macropore recovery process.

Our extensive experiments revealed that the tip material plays a determining role in triggering SM macropore recovery. Hard materials, like metals, graphite (pencil cores), and hard plastics (e.g., polystyrene), were found inefficient in generating colorful

patterns; while soft materials, such as low density polyethylene (LDPE) and polydimethylsiloxane (PDMS), were much easier in inducing a complete macropore recovery. Iridescent features with the same dimension as the writing tip immediately showed up following the movement of the tip. Figure 4-2A displays a greenish “UF” pattern written on a translucent macroporous SMP copolymer membrane templated from 300 nm silica microspheres. The typical cross-sectional scanning electron microscope (SEM) image in Figure 4-2B reveals that the macropores in the non- iridescent regions in Figure 4-2A are disordered and the surface of the deformed membrane is quite rough. By contrast, the macropores in the recovered iridescent regions are 3D highly ordered and the film surface is much smoother (Figure 4-2C). The average thickness of the macroporous layer changes from 2.77 ± 0.26 m for the disordered array to 4.56 ± 0.04 m for the recovered photonic crystal, indicating a 65% expansion of the deformed macropores. The different optical appearances of the translucent and the iridescent regions in Figure 4-2A can be quantitatively characterized by comparing their normal-incidence optical reflection spectra (Figure 4-2D). No apparent Bragg diffraction peaks are shown in the spectrum corresponding to the translucent region; while a distinct optical stop band located at ~ 543 nm with well- defined Fabry-Perot fringes is present in the spectrum obtained from the iridescent region. Importantly, the experimental spectrum matches well with the simulated spectrum using a scalar-wave approximation (SWA) model which assumes a perfect face-centered cubic (F.C.C.) crystalline lattice with its (111) planes normal to the incident light. This good match demonstrates the high crystalline quality of the writing- recovered photonic crystals. Moreover, the direct writing process is reversible. As

shown in the Supporting Information, the pre-written photonic crystal patterns can be entirely erased by drying the SMP membrane out of water. New photonic crystal features can then be written on the regenerated translucent film. This rewriting process can be repeated hundreds of times without apparent degradation in the chromogenic response of the SMP membranes.

In addition to macroscopic writing tools, atomic force microscopy was used to explore the capability in directly writing micro-/nano-scale photonic crystal features under well-controlled conditions. Figure 4-3A and 4-3B show AFM images of the designed “U” and “F” micropatterns written on a SMP copolymer membrane with 300 nm macropores using a 1 mm diameter sapphire spherical tip. Both letters were written with 140 µN contact force at a lateral writing speed of 5 µm/s. Each letter was written within a 100 × 100 µm2 region. As illustrated by the corresponding depth profiles in

Figure 4-3C and 4-3D, the letters protrude out from the rough membrane surface to a height of ~ 2 µm, and the minimum line width achieved by using the blunt tip is approximately 30 µm. The raised letters indicate that the SMP surface underwent a vertical transformation during the direct writing process, agreeing with the apparent thickness increase of the macroporous layer (~ 1.8 m) revealed by SEM (see Figure 4-

2B and 4-2C). Optical microscopy images (not shown here) illustrate that only the micropatterned areas reflect brilliant green light. A further observation of the SMP surface topography by higher resolution AFM imaging (Figure 4-4) show not only that the patterned areas are much smoother than the unpatterned areas (Figure 4-4C and 4-

4D), but also, the ordered arrangement of the macropores only appears on the patterned areas (Figure 4-4A and 4-4B). The root-mean-square (RMS) linear profile

roughness (Rq) of the patterned and unpatterned regions is 5.53 ± 0.75 µm and 50.51 ±

8.93 µm, respectively. The combination of the above observations including the ordered surface macropore structure supports that the reflective “UF” micropatterns are periodic arrays of recovered macropores.

To determine the critical contact pressure that can trigger the recovery of the deformed macropores during direct writing, a series of microscopic lines were written with decreasing force by controlling the set-point voltage applied to the AFM cantilever.

Figure 5A shows an optical microscope image (in transmission mode) of 6 lines written with 13.8, 27.7, 138, 277, 830, and 1380 N force (from left to right, corresponding to

0.005, 0.01, 0.05, 0.1, 0.3, and 0.5 V set-point voltage). The tip writing speed was held constant at 1 µm/s. In addition to the apparent difference in line width as revealed by the optical microscope image, other characteristics of the written lines were identified by

AFM images (Figure 4-5B-4-5G) and the corresponding depth profiles (Figure 4-5H-4-

5M). The line widths determined by both AFM and optical microscope images decrease from ~60 µm for the maximum force (1380 N) to ~ 25 m for the minimal force; while the heights of the protruding lines are nearly constant at ~ 1.2 µm. This means the SMP can recover to its permanent shape due to the tip-sample interaction, which is in part caused by the attractive force (adhesion) between the tip and the membrane. This attractive force is contributed by both the van der Waals interactions and the capillary force generated by the water meniscus bridging in between the tip-sample. The average

F ave maximum attractive force ( attr ) can be determined from the measurement of the pull- off force required to disengage the contact of the AFM tip with the sample. The contact force was calculated as the difference between the and the pulling force applied by

the tip. The minimum pressure that can induce the SM recovery was determined by

PFA / P F mm, where m is the minimum pressure, m is the minimum contact force and

2 Ar  is the contact area. The minimum contact radius r is calculated as half of the line width. Table 4-1 summarizes the writing forces, the resulting line widths, and the calculated contact pressures. The minimum contact pressure that can cause the macropore recovery was determined to be ~ 26 kPa. Quasi–static indentation was also explored to compare with the dynamic direct writing process. However, even with an applied force 1000 times larger than the writing one, the SMP surface was barely recovered by quasi-static indentation, as there was no distinguishable diffractive photonic crystal pattern generated. The SM recovery mechanisms and the difference between these two processes will be discussed later.

To further investigate the writing speed effects on the SMP surface recovery, as well as the minimal line width enabled by AFM directly writing, a 20 µm diameter borosilicate spherical tip was used to perform a series of writing experiments. By using a smaller tip radius, the resolution is significantly increased along with the sensitivity in writing speed. The AFM images in Figure 4-6 show nanoscopic lines written with the same force (6 µN), but the writing speed was increased from 0.2 µm/s (Figure 4-6A) to

20 µm/s (Figure 4-6D). The characteristics of the resulting nanopatterns including line widths and protruding heights are summarized in Table 4-2. It is apparent that both the line widths and heights of the recovered nanopatterns increased with higher writing speed. This set of experiments confirms that the SM recovery of the deformed macropores is dependent on the lateral motion and perturbation between the AFM tip and the SMP membrane.

All above experimental results have indicated that the direct writing approach is not a simple and straightforward extension of the static printing technology as reported in our previous work. From an energy perspective, the SM deformation and recovery processes are due to the energy transformation between the external (i.e., capillary pressure induced by water evaporation, applied contact force, and shear stress caused by tip lateral motion) and the internal (e.g., polymer chain movement, internal energy change, and stored elastic energy) of the SMP system. At room temperature, the

ETPTA 20-co-PEGDA 600 copolymer is in its rubbery state, above the glass transition temperature (Tg ~ 42 C). The polymer chains have high mobility and the polymer behaves like a soft (viscous) elastic material. The mechanism of energy transformation and macropore collapse during the water evaporation process has been explained in our previous work. The water evaporation process has two effects on the polymer structure: (1) the large surface tension of water collapses the originally ordered macropores; (2) there is energy dissipation during water evaporation that freezes polymer chain mobility. To reactivate the squeezed polymer chains and trigger the collapsed macropore recovery to its original configuration, external input energy is needed, or, equivalently, a reverse process to water evaporation is needed. Mechanical stress  in the form of either statically or dynamically applied force by a rubber stamp or a writing tip can input energy into the SMP system. In the case of direct writing, the energy required to overcome the SM activation barrier is provided by sliding the tip across the SMP surface with a compressive force. The kinetic energy of the tip is transferred to the polymer matrix in the form of shear deformation and vibration. The combination of shear and vibration, which increases the internal energy of the SMP

system, provides the energy to activate polymer chain mobility and trigger the macropore recovery. Our experimental results show that with higher tip sliding speed, a more complete recovery of SMP was achieved, which was presented as a higher recovered line width and height (see Figure 4-6). This is in accordance with intuition that the more input energy, the higher density of activated polymer chains.

It has been argued that the shear stress field can induce changes in the conformation of intermolecular bonds and polymer chain flow in glassy polymers at temperatures above Tg. In addition, a recent study was able to directly measure stress- induced molecular mobility in glassy polymers. Mobility was shown to increase by 10 

1000 folds after stress was applied. Furthermore, nanoscopically raised patterns were observed when a polyethyleneoxide (PEO) film was raster-scanned by an AFM tip at ambient conditions. Viscoelastic effects and localized heating caused by rupture of the adhesive bonds between the tip and the polymer, which could raise the local surface temperature by up to several hundred kelvin, were attributed to the unexpected formation of the raised areas during scanning. These studies support what we observed with the effects of the tip materials (e.g., LDPE vs. stainless stain tip) and the varying writing speed (i.e., strain rate) on the SM macropore recovery. LDPE-wrapped tips, which could form stronger bonds with the ETPTA 20-co-PEGDA 600 copolymers than stainless steel tips, are thus more efficient in inducing a more complete SM recovery during direct writing. In the case of quasi-static indentation, only vertical contact between the tip and the SMP membrane was involved. During indentation, the AFM tip compressed the macroporous structure to a more squeezed configuration. The majority of the external energy was stored in the elastic deformation of the polymer matrix. Only

very limited kinetic energy was transferred as internal energy to activate the polymer chains. As a result, a comparatively larger force is expected to reactivate the polymer chain mobility than in the dynamic writing case. In retraction, the stored elastic potential energy was gradually released as the SMP surface returned to its initial contact height.

Then the attractive adhesion force between the tip and the sample acted as the subsequent recovery force. Our experimental results support this conjecture. It is worthy to point out that the minimum force that can cause macropore recovery in quasi-static indentation is two to three orders magnitude higher than that in dynamic writing.

Assuming the indention process is one extreme case in writing for which the lateral speed is zero, then, it is clear that the dominant energy to induce SM recovery comes from the lateral movement of the writing tip.

Conclusion

In conclusion, by integrating the well-established templating nanofabrication with a new type of pressure-responsive SMP, we have developed a direct writing technology for fabricating arbitrary 3D ordered macroporous photonic crystal patterns in a single step. Both macroscopic and nanoscopic photonic crystal features can be reversibly patterned and erased, promising for making reconfigurable/rewritable nanooptical devices. Systematic experiments have revealed the importance of the material selection, dimension, applied force, and speed of the tips in affecting the SM recovery of

3D ordered macropores. Importantly, the dynamic writing approach exhibits significant differences in SM recovery mechanisms and critical recovery force than quasi-static indentation. Besides straightforward applications in photonic crystal devices and nanooptics, the striking chromogenic effects induced by the disorder-to-order transition during SM recovery of ordered macropores, the manifest protrusion of the recovered

regions, the sensitivity of the SMP membranes to various mechanical stresses, the unusual room-temperature operation for the entire shape-memory cycle, and the microscopic resolution of the directly written features could add new dimensions to many existing and future applications, such as in mechanochromic stress and impact sensors, rewritable high-density data storage, and tunable phononic crystals for controlling the flow of phonons.

Figure 4-1. Schematic illustration showing the direct writing of microscopic 3D photonic crystal patterns (letters “U” and “F”) on a macroporous SMP membrane with collapsed macropores using an AFM tip.

Figure 4-2. A green-colored, handwritten “UF” pattern on a translucent macroporous SMP membrane with collapsed 300 nm macropores. A) Photograph of a green-colored, handwritten “UF” pattern on a translucent macroporous SMP membrane with collapsed 300 nm macropores. B) Typical cross-sectional SEM image of the translucent region in A). C) Typical cross-sectional SEM image of the iridescent region in A). D) Normal-incidence optical reflection spectra obtained from the iridescent and the translucent regions of the sample in A). The simulated spectrum using a SWA model is also shown to compare with the experimental results. Photo courtesy of author.

Figure 4-3. A micropattern “U” directly written on a macroporous SMP membrane. A) AFM image of a micropattern “U” directly written on a macroporous SMP membrane using a 1 mm diameter sapphire spherical tip. B) AFM image of a micropattern “F”. C) Height profile scanned across the dashed line in A). D) Height profile scanned across the dashed line in B).

Figure 4-4. AFM image of a recovered area on the micropatterned letter “F”. A) Higher resolution AFM image of a recovered area on the micropatterned letter “F” in Figure 3b. B) Higher resolution AFM image of an unpatterned area in Figure 3b. C) Height profile scanned across the dashed line in A). D) Height profile scanned across the dashed line in B).

Figure 4-5. Microscope image of microscopic lines written with increasing forces. A) Optical microscope image of microscopic lines written with increasing forces from left to right. B-G) AFM images of the lines in (a). H-M) Height profiles scanned across the dashed lines in B)-G).

Figure 4-6. AFM images of nanoscopic lines written with different AFM tip speeds. A) 0.2 m/s. B) 1 m/s. C) 5 m/s. D) 20 m/s.

Figure 4-7. Photograph of a rubber stamp and an iridescent “A+” pattern printed on a translucent macroporous SMP membrane. A) Photograph of a rubber stamp with a “+A” relief pattern on its surface. B) Photograph of an iridescent “A+” pattern printed on a translucent macroporous SMP membrane templated from 280 nm silica microspheres. Photo courtesy of author.

Table 4-1. Dependence of the recovered line widths on the parameters of the nanoindentation-based direct writing process. Set-point 0.005 0.01 0.05 0.1 0.3 0.5 Voltage (V) Writing Force 13.8 27.7 138 277 830 1380 (N) Line Width 25.8  33.0  40.4  44.3  53.8  60.4  (m) 1.9 2.7 1.1 1.6 2.2 1.1 Contact Pressure 26.2 32.0 106.7 177.4 361.1 477.4 (kPa)

Table 4-2. Dependence of the recovered line width and height on the writing speed of the AFM tip. Writing Speed 0.2 1 5 20 (m/s) Line Width 4.1  0.4 5.6  0.9 6.0  0.9 6.4  0.4 (m) Line Height 301  12 407  17 477  20 526  29 (nm)

CHAPTER 5 OPRICALLY BISTABLE MACROPOROUS PHOTONIC CRYSTALS ENABLED BY THERMORESPONSIVE SHAPE MEMORY POLYMERS

Introduction

Photonic crystals are periodic dielectric structures with a forbidden band gap for electromagnetic waves, analogous to the electronic band gap in semiconductors.

Photons with energies lying in this photonic band gap (PBG) cannot propagate through the dielectric medium, providing vast opportunities in controlling the flow of light in miniature volumes for a large variety of applications ranging from all-optical integrated circuits to quantum information processing.46,78,104,141-143 Compared with conventional photonic crystals with fixed PBGs, which are useful for developing passive nanooptical devices, tunable photonic crystals with adjustable PBGs have recently attracted great research interest as they could enable active devices, like low-threshold tunable lasers, full-color displays, optical switches, and chemical sensors.48,85,105,108,110,112,113,144,145 A large variety of tuning mechanisms, such as mechanical pressure, temperature variation, electrical and magnetic fields, solvent swelling, and redox reactions have been exploited to change either the photonic crystal structural parameters (e.g., lattice constant and crystalline structure) or the effective refractive index of the diffractive media.85,146-155 For instance, mechanical pressure is a commonly used driving force for tuning the crystalline lattice spacing of soft photonic crystals consisting of elastic materials (e.g., elastomers and gels). Externally applied magnetic fields have been extensively explored in controlling the assembled structures of magnetic nanoparticles to achieve tunable PBGs. Refractive index tuning is also widely utilized in altering structural colors.154 However, traditional tunable photonic crystals usually cannot memorize the temporarily configured optical microstructures during the tuning process

and they rapidly return to the original photonic crystal structures once the external driving forces are removed.

Photonic crystals with optically bistable states, which complement currently available tunable photonic crystals, could provide a unique opportunity in realizing reconfigurable and rewritable optical devices. For example, in sharp contrast with traditional subtractive microfabrication (e.g., photolithography followed by reactive ion etching), functional optical microstructures can be patterned, erased, and rewritten on these optically bistable photonic crystals to accommodate different applications needs.

This could dramatically reduce the complexity and fabrication cost of developing a large number of application-specific photonic crystal devices.56,108,109,156,157 Shape memory polymers (SMPs), which can memorize and recover their permanent shapes in response to various external stimuli (e.g., heat, solvent, light, electrical and magnetic fields), may hold the key to reconfigurable photonic crystals with bistable states. They have also been widely used in a spectrum of applications ranging from smart surgical stents and sutures to aerospace morphing structures.29,38,84,122,158,159 Shape memory

(SM) for thermoresponsive SMPs is usually achieved in three steps including programming, storage, and recovery. In SM programming, a SMP sample in its permanent shape is heated above a specific transition temperature (Tm), such as the polymer glass transition temperature (Tg), and then deformed to a temporary shape.

After cooling below Tm, the sample can store the “frozen” temporary shape at ambient conditions for a long period of time. Finally, the memorized permanent shape can be recovered when the sample is reheated above Tm, and entropy elasticity of SMPs accounts for the shape restoration.1,18,160,161

Although the mechanically stable permanent and temporary shapes of SMPs could enable reconfigurable photonic crystals with optically bistable states, smart SMPs have rarely been used in previous photonic crystal studies. Programmable and shape- memorizing micro-optical devices, such as two-dimensional (2-D) microlens and microprism arrays, diffraction gratings, and holograms, have been demonstrated by using a semi-crystalline shape memory elastomer  crosslinked poly(ethylene-co-vinyl acetate). Similar thermoresponsive shape-memory 2-D photonic crystals consisting of periodic microbowls (templated from 870 nm polystyrene microspheres) in a polydiolcitrates-based elastomer have recently been demonstrated as active diffractive elements for various optical applications.139,140,162-165 Three-dimensionally (3-D) ordered elastomeric polymer opal films, which exhibit tunable and reversible PBGs triggered by light, heat, and mechanical strains, have been assembled by using specifically engineered core-interlayer-shell (CIS) polymer microspheres.48,85,105-113,144,145

Unfortunately, the SMP-enabled tunable photonic crystals in these previous studies suffer from either low PBG (indeed optical stop band) amplitudes caused by the 2-D nature of the surface gratings or limited material selection.

Methods

Templating Fabrication of Thermoresponsive Macroporous SMP Photonic Crystal Membranes:

Monodispersed silica microspheres with 300 nm diameter were synthesized by the standard Stöber method.131 The resulting silica particles with diameter standard deviation less than 5% were purified in 200-proof ethanol by multiple centrifugation and redispersion cycles (at least 5 times). The purified silica microspheres were then self- assembled on a glass microslide by the convective self-assembly technology to form

close-packed multilayer colloidal crystals. The crystal thickness was controlled to ~ 5

m (or ~ 20 colloidal monolayers) by adjusting the particle volume fraction of the silica microsphere-ethanol suspension. A double-sided adhesive tape (~ 1.7 mm thick) was used as a spacer to separate the microslide with the assembled colloidal crystal on its surface and a bare glass microslide. The viscous oligomer mixture (CN745A70,

Sartomer) consisting of trifunctional acrylated urethane and tripropylene glycol diacrylated was pre-heated at 90 °C for 5 min. Darocur 1173 (2-hydroxy-2-methyl-1- phenyl-1-propanone, BASF, 0.8 wt %) was added as the photoinitiator. The oligomer mixture was injected in the above microslide sandwich cell. The replacement of air inbetween silica microspheres with the index-matching oligomer mixture made the cell transparent. The sample was transferred to a pulsed UV curing system (RC 742,

Xenon) and the oligomer mixture was rapidly polymerized by exposure to UV radiation for 4 s. The polymerized film was immersed in a 1 vol % hydrofluoric acid aqueous solution for 48 h and then rinsed with deionized water. After blow-drying with compressed air, the final self-standing macroporous PU-co-TPGDA copolymer membrane showed striking iridescent colors.

Heat-Induced SM Programming and Recovery:

The templated macroporous SMP copolymer membrane was placed inbetween a pre-cleaned glass substrate and a smaller glass piece. The sample was then held by two stainless steel plates and a clamp force of 200 lb was applied using a manual hydraulic press (Carver Model C). The temperature of the system rapidly increased from room temperature to ~ 90 °C and stayed at this temperature for 3 min. The sample then cooled down to room temperature. After releasing the clamp force, the pressed region

100

became transparent. The deformed sample was finally placed in an oven preset at 90

C for 2 min to recover the deformed macropores.

Sample Characterization:

Nanosensor PPP-NCHR probe (tip radius < 10 nm). In-situ nanoindentation tests were conducted using a MFP-3D NanoIndenter (Asylum Research) with a spherical sapphire indenter (tip radius ~ 125 µm). Such configuration of the instrument has a force and displacement resolution less than 3 µN and 1 nm, respectively. Differential scanning calorimetric measurements were performed from 0 to 200 C at a heating rate of 10 C min-1 using a TA Instruments DSC Q1000 and an empty pan as reference. The apparent water contact angle was measured using a goniometer (NRL C.A.

Goniometer, Ramé-Hart Inc.) with autopipetting and imaging systems. Normal-incidence optical reflection spectra were taken using an Ocean Optics HR4000 high-resolution fiber optic vis-NIR spectrometer with a reflection probe (R600-7) and a tungsten halogen light source (LS-1). Absolute reflectivity was obtained as the ratio of the sample spectrum and a reference spectrum, which was the optical density obtained from an aluminum-sputtered (1000 nm thickness) silicon wafer.

Scalar Wave Approximation Optical Modeling:

The scalar wave theory implemented for periodic dielectric structures was utilized in modeling the normal-incidence optical reflection spectra from templated macroporous

SMP photonic crystals. In this theory, Maxwell’s equations are solved for a periodic

101

dielectric assuming that one may neglect diffraction from all but one set of crystalline planes (e.g., (111) planes in this work). No adjustable parameters are used in SWA modeling, since the size of the macropores and the crystal thickness can be independently determined from SEM images, and the refractive index of the copolymer is known.

Results

We report a generalized templating approach for fabricating thermoresponsive 3-

D macroporous SMP photonic crystals with optically bistable states. The technology can be easily applied to nearly all functional SMPs, provided they can survive the template removal process, such as a brief hydrofluoric acid wash for removing templating silica colloidal crystals.97In addition, the templated macroporous SMP photonic crystals with high crystalline qualities exhibit strong diffractive effects. Moreover, the heat-triggered transition between a disordered temporary state and a 3-D ordered permanent state leads to an easily perceived color change and a significant difference in their optical responses. Furthermore, the 3-D photonic crystal structure provides a simple and sensitive optical technology for quantitatively characterize the intriguing shape memory effects at nanoscale.

The thermoresponsive macroporous SMP photonic crystal membranes with 3-D ordered macropores were prepared by templating fabrication using self-assembled silica colloidal crystals as structural templates. The convective self-assembly technology was utilized in organizing monodispersed silica microspheres with 300 nm diameter into highly ordered colloidal single crystals.76,97 The crystal thickness was controlled to ~ 5

m by adjusting the particle volume fraction of the silica microspheres/ethanol

102

suspension to ~ 1.0 %. Through capillary interactions, the interstitial air in between the close-packed silica microspheres was replaced by a commercial oligomer mixture

(CN745A70, Sartomer, viscosity 2200 cps at 60 C, refractive index 1.4795) consisting of trifunctional acrylated urethane and tripropylene glycol diacrylated (TPGDA). The oligomer mixture was then photopolymerized by exposing to UV radiation. We termed the resulting polyester/polyether copolymer as PU-co-TPGDA in this paper. The templating silica microspheres were finally removed by wet-etching in a 1 vol % hydrofluoric acid aqueous solution. The final self-standing macroporous SMP membranes are rigid and show brilliant diffractive colors which depend on the size of the macropores and the viewing angle. In this paper, we used the commercial PU-co-

TPGDA copolymer as a model thermoresponsive SMP, though the technology can be easily applied to many other types of SMPs.

The glass transition temperature of the PU-co-TPGDA copolymer, which is a critical parameter in heat-induced SM programming and recovery processes, was evaluated by differential scanning calorimetry (DSC). The typical DSC plot of a macroporous copolymer memberane in Figure 5-1A shows a single Tg of ~ 85.6 C, indicating the crosslinked copolymer is a homogeneous mixture of the two components.

In addition, no apparent crystallization dips show up in the DSC plot for temperature from 0 to 200 C (see the complete DSC plot with both heating and cooling cycles in

Figure 5-8), confirming the copolymer is amorphous in this temperature range. The

Young’s modulus of the templated macroporous copolymer membrane was characterized by in-situ nanoindentation tests. Figure 5-1B compares the average moduli of 3 membranes with 300 nm macropores indented by applying 100 mN force.

103

More than 30 points on each sample were measured to obtain the average modulus and its standard deviation. All 3 macroporous samples possess Young’s modulus of ~ 3

GPa, which is higher than the modulus measured by conventional tensile tests (~ 1.4

GPa with tensile strength of ~ 110 MPa and yielding strain of ~ 0.08, provided by the vendor). The typical water drop profile in Figure 5-1C shows the macroporous PU-co-

TPGDA copolymer is nearly hydrophobic with an apparent water contact angle (CA) of ~

81.

The high Tg and mechanical strength, as well as the hydrophobic properties, make the PU-co-TPGDA copolymer a good candidate for making reconfigurable/rewritable photonic crystals with high durability and environmental stability. The schematic illustrations in Figure 5-2 show the microstructural transitions during heat-induced SM programming and recovery processes of a macroporous PU- co-TPGDA photonic crystal membrane. In programming, the original, 3-D highly ordered macropores (permanent state) are deformed into disordered arrays when the copolymer sample is heated above its Tg, while simultaneously a sufficient pressure is applied on the softened copolymer. The deformed sample is then cooled below Tg to fix the temporary shape (disordered array) which can be stored at room temperature for a long period of time. Shape memory recovery to the permanent, 3-D ordered structure occurs when the deformed sample is reheated to above Tg. Similar to conventional thermoresponsive SMPs, entropy elasticity is believed to be the major reason for the heat-triggered macropore recovery.

The photographs and the corresponding cross-sectional scanning electron microscope (SEM) images in Figure 5-3 show the color and microstructure changes of a

104

macroporous PU-co-TPGDA membrane with 300 nm macropores transited from the permanent state (Figure 5-3A and 5-3B) to the deformed state (Figure 5-3C and 5-3D), and then to the recovered state (Figure 5-3E and 5-3F). The original macroporous copolymer film exhibits a shining greenish color (Figure 5-3A) which is caused by Bragg diffraction of visible light from the 3-D highly ordered macroporous photonic crystal (see the cross-sectional SEM image in Figure 5-3B). The diffractive color disappeared

(Figure 5-3C) when the sample was heated to 90 C, while a clamp force of 200 lb was applied through a rectangle-shaped glass piece (~ 1.5 × 3.7 cm) using a manual hydraulic press. After cooling down the sample to room temperature and releasing the clamp force, the deformed region became nearly transparent. The loss of the periodicity of the original photonic crystal structure and the deformation of the spherical macropores, which can be clearly seen from the cross-sectional SEM image in Figure 5-

3D, attributed to the discoloration of the membrane. In addition, the average thickness of the macroporous layer significantly reduced from ~ 5.2 m to ~ 3.2 m in the above heat-induced programming process. This abrupt change in film thickness led to the clear edges between the colorful and the transparent regions in Figure 5-3C. The retainment of the greenish color around the upper right corner of the rectangular box was caused by a defect in the stainless steel chamber of the hydraulic press which affected the even distribution of pressure applied on the macroporous sample.

The disordered array of collapsed macropores can be recovered back to the permanent photonic crystal structure by reheating the deformed sample above Tg of the

PU-co-TPGDA copolymer. Figure 5-3E shows the same sample as Figure 5-3C after heating the membrane in an oven at 90 C for 2 min. The recovery of the vivid greenish

105

color of the deformed rectangular region is clearly evident. The cross-sectional SEM image in Figure 5-3F confirms the restoration of the ordered photonic crystal structure after the above thermal treatment. However, our extensive SEM characterization indicated that the heat-triggered macropore recovery is not fully complete (i.e., the strain recovery rate is not 100%). The average thickness of the recovered macroporous layer in Figure 5-3F is ~ 4.9 m, which is about 94% of the thickness of the original macroporous photonic crystal. Additionally, by comparing the SEM images in Figure 5-

3B and 5-3F, it is apparent that the surface of the restored macroporous layer is not as smooth as the original sample. This is another evidence that indicates the incomplete macropore recovery. To quantitatively characterize the surface microsturctures of the macroporous photonic crystals in the heat-induced SM programming and recovery processes, we conudcted extensive atomic force microscopy (AFM) imaging. Figure 5-4 presents AFM topographic images and the corresponding height profiles scanned across the dashed lines for the same macroporous samples as shown in Figure 5-3.

The surface of the original photonic crystal is smooth and the templated macropores are highly ordered (see Figure 5-4A and 5-4B). By contrast, the same sample in the deformed state (Figure 5-4C and 5-4D) has a much rougher surface and the ordered macroporous structure is entirely lost, silimar to the cross-sectional SEM image as shown in Figure 5-3D. The partial recovery of the original smooth surface and the long- range ordering of the macropores is apparently shown by the recovered sample (Figure

5-4E and 5-4F). The significant changes in surface roughness can be quantitatively characterized by the root mean square (RMS) roughness, Rq, using AFM topographic images. The average Rq is 8.8 ± 0.7 nm, 47.9 ± 16.2 nm, and 14.0 ± 1.9 nm,

106

corresponding to the above 3 samples with the permanent, deformed, and recovered state, respectively. The surface of the deformed sample is substantially rougher than the original and the recovered samples.

The microstructural transition associated with the heat-induced SM programming and recovery processes, which induces easily perceived color changes, can also be quantitatively characterized by normal-incidence optical reflection measurements.

Figure 5-5 compares the reflection spectra obtained from the same macroporous SMP sample with 300 nm macropores in the permanent state (red curve), deformed state

(black curve), and recovered state (blue curve). The original macroporous photonic crystal with 3-D highly ordered macropores (see Figure 5-3B) exhibits a strong PBG peak at ~ 539 nm with well-defined Fabry-Perot fringes. The PBG peak position agrees with the calculation using the Bragg diffraction equation:

max 2 ndeff   sin (5-1)

n where eff is the effective refractive index of the macroporous photonic crystal (~ 1.12 by assuming the refractive index of the SMP copolymer is 1.47 and the volume fraction of the copolymer and air is 0.26 and 0.74, respectively), d is the inter-plane distance (

300 2 3 o nm), and   90 for normal incidence. Moreover, the experimental reflection spectrum of the macroporous sample in the permanent state matches well with the theoretical spectrum (red curve) simulated using a scalar wave approximation (SWA) model, which assumes a perfect face-centered cubic (f.c.c.) structure with the (111) crystalline planes parallel to the sample surface.

In sharp contrast with the distinct PBG peak displayed by the original macroporous SMP photonic crystal, the deformed sample with disordered macropores

107

(see Figure 5-3D) shows no apparent PBG peaks in the normal-incidence reflection spectrum. This complete discoloration represents a major difference between traditional tunable photonic crystals and the new SMP-enabled photonic crystals. For traditional tunable photonic crystals, the PBG peaks shift to different locations depending on the changes in the lattice spacing and/or effective refractive index. The wavelength shift is

 ~100 usually small ( max nm) and the PBG peak returns to the original position once the external stress is released. This means the temporarily deformed photonic crystal structures cannot be memorized. By contrast, the new SMP-enabled photonic crystals exhibit two optically stable states corresponding to the permanent (ordered) and the temporary (disordered) states of the macropores. This manifest order-disorder transition

  leads to the easily perceived color change and a large max (indeed it equals to max of the original photonic crystal as no PBG exists in the deformed state). When the deformed macropores recovered back to the 3-D ordered photonic crystal structure by reheating up the sample above Tg of the copolymer, the PBG peak and the Fabry-Perot fringes reappeared in the reflection spectrum. However, similar to the above SEM and

AFM results, the amplitude of the PBG peak is only partially recovered. In addition, the peak position slightly blue-shifts to ~ 530 nm, agreeing with the small decrease in the macroporous layer thickness as observed by SEM imaging.

Above, we have shown that optical reflection measurements provide a straightforward and sensitive methodology in characterizing microscopic structural changes of macroporous photonic crystals associated with heat-induced SM programming and recovery. This simple optical technology also enables in-situ characterization of an important shape memory parameter  the SM recovery response

108

speed of the thermoresponsive PU-co-TPGDA copolymer. Figure 5-6A shows the stacked optical reflection spectra obtained from a macroporous SMP sample which was placed on a hot plate preset at 90 ± 1 C. The optical spectrometer was set to automatically collect a whole spectrum for every second. The color-coded reflection amplitude gradually increases with longer heating duration and it reaches a plateau after ~ 200 s. This can be easily seen from the 2-D representation of the time-resolved reflection spectra in Figure 5-6B. Considering slow heat conduction through a thick, solid copolymer backing layer (~ 1700 m thick) to the thin macroporous layer (~ 5 m thick), as well as inevitable heat dissipation into the surrounding environment (at room temperature), it will take a while for the top SMP photonic crystal layer to be above its

Tg. Therefore, we believe the real SM recovery response speed at 90 C should be much less than 200 s. Indeed, the SM recovery speed is a sensitive function of the reheating temperature. Our experiment showed that a deformed macroporous SMP membrane recovered its vivid diffractive color within 3 s when directly heated by a heat gun.

The thermoresponsive SMP photonic crystals can be reused for at least several tens of times without apparent degradation in their optical performance. Figure 7 shows the absolute reflection amplitudes at 525 nm and the corresponding photographs of a macroporous PU-co-TPGDA copolymer membrane (O1) being cyclically deformed (D1,

D2, D3, D4, D5) and recovered (R1, R2, R3, R4, R5). The deformed and the recovered regions exhibit similar colors and absolute reflection amplitudes, indicating the high reproducibility of the heat-induced SM programming and recovery processes. This

109

recyclability is critically important for developing reconfigurable/rewritable photonic crystal devices.

Conclusion

In conclusion, we have developed a universal templating nanofabrication technology for making 3-D macroporous SMP photonic crystals with optically bistable states. The heat-induced transition between an ordered permanent state and a disordered temporary state leads to tremendous changes in the appearance and the diffractive properties of the photonic crystals. The high reproducibility in optical responses during multiple SM programming-recovery cycles, high Tg and mechanical strength, and hydrophobicity of the model PU-co-TPGDA copolymer demonstrate its merits in making reconfigurable/rewritable photonic crystal devices. Importantly, the thin macroporous photonic crystal structure enables a simple, quantitative, and sensitive optical technology for investigating the intriguing nanoscopic SM effects. In addition to reconfigurable nanooptical devices, the thermoresponsive macroporous SMP membranes with adjustable open or closed macropores could find many other applications, such as hydrophobic coatings with programmable wettability, smart membranes for size-exclusive filtration, and light-regulating coatings for energy-efficient buildings.

110

Figure 5-1. PU-co-TPGDA copolymer. (A) Typical DSC plot of PU-co-TPGDA copolymer. (B) Young’s modulus of 3 macroporous PU-co-TPGDA copolymer membranes indented with the same force (100 mN). (C) Water drop profile on a freshly prepared macroporous PU-co-TPGDA membrane with 300 nm macropores. Photo courtesy of author.

111

Figure 5-2. Schematic illustration showing the heat-induced programming and recovery steps of a thermoresponsive macroporous SMP photonic crystal membrane. (A) Permanent state with 3-D ordered macropores. (B) Deformed state with collapsed macropores. (C) Recovered state with reopened macropores.

112

Figure 5-3. Photographs and SEM images showing the apparent changes in reflective colors and microstructures during heat-induced programming and recovery steps of a macroporous PU-co-TPGDA copolymer membrane with 300 nm macropores. (A, B) Permanent state with 3-D ordered macropores. (C, D) Deformed state with collapsed macropores. (E, F) Thermally recovered state with reopened macropores. Photo courtesy of author.

113

Figure 5-4. AFM images and the corresponding height profiles (across the dashed lines) showing the changes in surface microstructures during heat-induced programming and recovery steps of a macroporous PU-co-TPGDA copolymer membrane with 300 nm macropores. (A, B) Permanent state with ordered macropores. (C, D) Deformed state with collapsed macropores. (E, F) Thermally recovered state with reopened macropores.

114

100 Permanent state Deformed state 80 Recovered state SWA simulation

Reflection(%)

0 400 450 500 550 600 650 700 750 800 Wavelength(nm)

Figure 5-5. Normal-incidence optical reflection spectra showing the permanent, deformed, and thermally recovered states of a macroporous PU-co-TPGDA copolymer membrane with 300 nm macropores. The simulated spectrum using a SWA model is also shown to compare with the experimental measurements.

115

Figure 5-6. Time-resolved normal-incidence optical reflection spectra during thermally induced recovery of a macroporous PU-co-TPGDA copolymer membrane with collapsed 300 nm macropores. (A) 3-D plot. (B) Color-coded 2-D plot.

116

100

Reflection (%) 20

O1 D1 R1 D2 R2 D3 R3 D4 R4 D5 R5

Figure 5-7. Reflection amplitudes at 525 nm wavelength and the corresponding photographs of a macroporous PU-co-TPGDA copolymer membrane with 300 nm macropores cyclically deformed (D1, D2, D3, D4, D5) and recovered (R1, R2, R3, R4, R5). O1 indicates the original sample. Photo courtesy of author.

117

Figure 5-8. Complete DSC plot of PU-co-TPGDA copolymer showing both heating and cooling cycles.

118

CHAPTER 6

INSPIRED BY SCALABLE COLLOIDAL TEMPLATING TECHNOLOGY AND POROUS COATING THAT ENABLES THE FABRICATION OF BROADBAND TUNABLE AR COATING.

Introduction

Shape memory materials (SMPs) with sophisticated compliable performance evolve tremendous desirable. Endowing intelligence to materials and programming materials on human desiring seems to be an interminable task. By integrating the intrinsic behavior of the materials and external stimuli, the elastic polymers exhibit a unique memorized functionality in response to variation of circumstance.37,38,64,93,166 The practical potential applications of (SMP) have been unprecedented in a number of different spheres, for instance, biomedical and optical tunable device, shape-shifting morphing wings, self-folding origami.112,159,167 Although current efforts to design reversible shape memory materials have been reported in many areas, the competing responsive time still suffer in controversies. In particular, increasing the responsive time of SMPs and leveraging the shape memory effect at nanoscale is virtually unexplored.

On the other hand, conquering the light reflection at the surface of the lens, digital display, aircraft-detector, and solar-panel, etal plays such a crucial role in the recent decade. The irritation of light-reflection dramatically responded in many fields, such as minimal resolution of photolithography, absorbing efficiency of solar panel, acuteness of optical lens, and distinguishability of space telescope. Most prevalent method to reduce light reflection is adding one low refractive-index materials as a buffer layer between the air and coating substrate according to the scalar diffraction theory.168-175 The reflection light will be deteriorated when the two wave vector generated by the two reflective light sits on the opposite phase. Lying on the principle of Fresnel formulas, the reflectivity

119

layer is dominated by the refractive index ratio on the above media and coating substrate. Thus, the antireflection criterion is calculated as

1/2 n1=(n0n2) ( 6-1) where n1, n0 and n2 are the refractive indices of the antireflection layer. Media from the incident side (normally is air) and a substrate. The minimum reflection is also determined by the optimal effective coating thickness which is gratified by the quarter

176-180 wavelength design (ncdc= λ0/4) and Fresnel reflection. The periodic non-close- packed hexagonal nanoarrays are observed in Nocturnal moths-eye.171,181 By integrating the bottom-up, fast-grown crystal and biomimic nanostructure via nature- inspired has emerged a new climax and made an unprecedented breakthrough.

Patterning 2D periodical structure and fine-tune, the intrinsic photonic band gap acts an essential attention as a powerful way to achieve broadband antireflection coating. After this promising approach, the most prevalent method to precisely manipulate the effective refractive index and coating thickness is the porous coating, because the porous truss can regulate light propagation through simply adjusting the pore diameter and arrangement. Thus, the single layer or multilayer porous coating is the widespread candidates by accurately amending the effective refractive index to achieve the adjustable band of ARCs.59,172,181 On the other hand, the proof-concept of a tunable antireflection coating is inevitable pervaded in current research. Oblique incidence the elastic materials, pattern replicated by soft materials, nanoporous polymers, self- assembly of block copolymers have been employed for fabricating effective optical layer with tunable antireflection property.173,176,182,183 However, most of the current

120

broadband antireflection coating are impeded by their complex triggering condition and inaccurately controlled at the ambient condition.

Methods

Templating Nanofabrication of Monolayer Macroporous 2D Photonic Crystal Membranes

Monodispersed silica spheres with diameter ranging from 100nm to 140nm were synthesized by the standard Stöber method.131 The as-synthesized silica particles were first purified in 200-proof ethanol by multiple centrifugation and re-dispersion cycles for

6 times and re-dispersed in ethylene glycol (Fisher >= 99.8%) with 2% particle volume fraction. The silica/ethylene glycol suspension was added dropwise to the surface of deionized water contained in a glass crystallizing dish. All the glass substrates and dishes were cleaned in Piranha solution (4:1 H2SO4: H2O) to remove organic residue.

The silica spheres were accumulated and form a colloidal monolayer on the air/water interface. The colloidal monolayers were transferred to the preimmersed glass microslide (75 mm × 25 mm) by using a clamp attached to a syringe pump (KD

Scientific 780-230) with vertically withdraw rate at 12.5 mm∕min. Next, we used two coated monolayer colloidal crystal glass substrates as template, which were stick together by a double-sided adhesive tape as a spacer with controlled space of 1 mm.

The gap (1mm) between two colloidal templates was filled with viscous oligomer mixture consisting of ethoxylated (20) trimethylolpropane triacrylate (SR415, Sartomer, MW

1176, viscosity 225 cps at 25 C, refractive index 1.470) and polyethylene glycol (600) diacrylate (SR610, Sartomer, MW 742, viscosity 90 cps at 25 C, refractive index 1.468) oligomers with varying volumetric ratios from 1:1 to 1:6. Darocur 1173 (2-hydroxy-2- methyl-1-phenyl-1-propanone, BASF, 1 wt %) was added as the photoinitiator. Since

121

the viscosity of oligomer mixture is low, the interstitial air in-between silica microspheres has been frothed in viscous oligomer mixtures by the capillary forth. The sandwiched sample with oligomer mixture was rapidly polymerized by exposure to UV radiation for 4 s with a pulsed UV curing system (RC 742, Xenon). Then, the polymerized film was soaked in a 5 vol % hydrofluoric acid aqueous solution for 1.5 h and rinsed by deionized water. After drying with kimwipe, free-standing monolayer macroporous SMP membranes of double side were fabricated.

Macroporous SMP Shows Tunability of Antireflection Property Which Accurately Controlled by Morphology of Macropores

To investigate the tunable antireflection property by controlling morphology of macropores through varying the peripheral circumstance, we prepared a 2 x 2 cm2 monolayer macroporous SMP film of double side as our tested sample. First, we rinsed the dry macroporous film by ethanol solution using damp Q-tip. Then, the ethanol- treated film showed outstanding antireflection property and proved by the transmission spectra after drying out of ethanol. For comparison, the film which has been treated by ethanol rinsed by deionized water again using damp Q-tip. As predicted, the antireflection property of water-treated film is apparently reducing which shows strong glare and expresses on the transmission spectra. Through considerable investigation, the tunable antireflection property can be repeated more than 100 times under variation of external circumstance.

Macroporous SMP Shows Tenability of Antireflection Property by Contacting Pressure

To investigate the extra-induced mechanism, we prepared a small piece of

PDMS with area (0.5x0.5 cm2). Then, we chose a single layer (100nm) macroporous membrane with two side coating rinsed by DI water. The macropores were collapsed

122

attributed to the strong capillary force during the evaporation of water, thus the macroporous membrane shows intensely glare effect due to the collapsed macropores.

To do the substantial comparison, we applied a contact pressure by a small piece of

PDMS on both side of the SMP membrane which has collapsed macropores.

Surprisingly, the glare has disappeared and instead by observable antireflection property. In addition to appropriate complex pattern, we use a stenciled leaf-pattern

(3cm x2.5cm). After being directly printed on the tunable antireflection SMP film, the leaf pattern shows a vivid leaf with high contrast of antireflection area and glare area. Then, we reshaped a piece of PDMS (4cmx1.5cm) as a chess pad and applied a small force between “chess pad” PDMS and SMP membrane. The printed film shows clearly “chess pad” with strong contrast in glare region and antireflection region as indicated in Figure

6-9. All the membrane can be reused for hundreds of times.

Sample Characterization

NCHR probe (tip radius < 10 nm). All AFM images were processed using the Scanning

Probe Imaging Processor (SPIP, Image Metrology Inc., Horsholm, Denmark) software.

Normal-incidence optical reflection spectra were obtained using an Ocean Optics

HR4000 high-resolution fiber optic vis-NIR spectrometer with a transmission probe

(QP400-2-SR) and a tungsten halogen light source (LS-1).

Results

We developed a novel reconfigurable antireflection elastic film created by templated macroporous shape memory polymer. Inspired by the microstructure of

123

miraculous moth-eye in nature, we integrate fast assembling 2D photonic crystals and shape memory polymers. We develop a novel stimuli-responsive shape memory polymer that deploys a promising tunable antireflection property with absolute control over the morphology of macrostructure in order to tailor the optical property. The performance of adjustable antireflection property can be predictively achieved by regulating the morphology of macropores. We show that the structurally designed film demonstrate the variation of antireflection property through 94.5% to 99.8% by selectively modified macropores film through drying out of water, contact pressure or ethanol. We further demonstrate an unusual cold “programming” that lead to the macropores deformed by capillary pressure in the process of drying out water. With the aid of tactile pressure or wiping by ethanol solvent, the deformed macroporous structure is snapped back to the freestanding state with the striking variation of antireflection property.

Fabrication of Tunable Antireflection Shape Memory Films

The large area and single layer colloidal template with close-packed microsphere as indicated in figure 6-8 A, B were fabricated using Langmuir-Blodgett method. The purified silica microspheres with the controlled diameter from 100-140 nm were uniformly dispersed into ethylene glycol with varying volumetric ratio from 2:98 to 5:95.

The microslide was fixed on the syringe pump on one side, at the same time, immersing the other side into the deionized water. Then, the silica/ethylene suspension was added dropwise to the surface of high purity deionized water contained by a glass tank until it was accumulated into a thin layer. The arm of syringe pump was lifted up with a constant speed at 12.55mm/min to ensure the floated silica microsphere are homogeneous and close-packed adhesive on the microslice implemented by the strong

124

capillary force of water. Then, two pieces of colloidal template were placed so that they confronted each other and assembled themselves as a sandwich structure. The ethoxylated (20) trimethylolpropane triacrylate (ETPTA20, MW1176, viscosity 225 cps at 25ºC, Tg ≈─ 40ºC, refractive index ≈1.470)co-polyethylene glycol(600) diacrylate

(PEGDA 600,MW742, viscosity 225cps at 25ºC, Tg ≈ ─ 42 ºC, refractive index ≈1.468) liquid oligomer was penetrated to the colloidal template driven by capillary pressure and substituted the interstitial between the close-packed microspheres. The combined copolymer was rapid photocured under the UV light at ambient condition. The embedded silica particles were etched out when the oligomer composite immersed in a dilute hydrofluoric acid aqueous solution. After removing the silica microspheres, the reversed structure performed a free-standing macroporous ETPTA-co-PEGAD copolymer film on each side. The intermost truss is then removed by HF. The final 2D self-supporting polymer nanolattice comprises a network of microvoids sitting on the polymer matrix. Consequently, the monolayer macroporous film demonstrates a unique tunable antireflection property at bistable state when the film is suffering the variable stimuli.

Reconfigurable Antireflection 2D Photonic Crystals Enabled by Pressure- Responsive Shape Memory Polymer

The reversible antireflection coating has been released for many years.

PAH/PAA reversible nanoporous thin films and amiphi-RAFT (reversible addition fragmentation transfer) polymer core-shell nanoparticles.174,184,185 In our study, we discovered that our macroporous film expressed the reversible antireflection property at the bistable state undergoing the different external stimuli. The effective reflective index can be fine-tuned in the versatile circumstance. Figure 6-1 schematically compares the

125

variation of micropores in the transition process, during which the opacity of the film significantly varies through the change of the micropores. Specifically, we defined the state that the micropores are fully opened as the permanent shape as dictated in figure

6-1A. In the structural porous coating, the 76 % volume fraction of hierarchical microvoids are occupied by air and the rest 24% volume fraction of the skeleton are employed by the SMP. Attributed to the unique structure which has a high volume fraction of air cavities, the effective refractive index subtlety decline. The porous coating forms graded index variation between air (refractive index =1), porous layer (refractive index ≈1.11) and bulk polymer (refractive index ≈1.47). The transmitted light film exhibit respectable AR property because of the destructive interference occurred.

On the other hand, owe to the strong capillary force during the water evaporation, the macropores can be squeezed and distorted due to the low polymer glass transition temperature (Tg) and compliant elastically property comparing to the strong capillary force of water at the ambient conditions that is defined as temporary shape shown in figure 6-1B. The deformed microarrays cannot break the interference of the reflection light, hence, it drastically reduces the AR property and shows appear glare-effect.

Promisingly, the deformed micropores can immediately snap back to the permanent shape that is defined as recovery shape as shown in figure 6-1C that has highly ordered pores by exposing to the external stimuli. Simultaneously, the antireflection property is fully recovered.

To illustrate the contact pressure-induced shape memory recovery, the engraved square and leaf pattern is exemplified by the photograph in figure 6-2A, B, that is triggered by the contact pressure between stencil template and deformed SMP

126

membrane directly. The clarify contour of leaf region can look through the membrane and shows clear “UF” caption in circumstance. The cross-section SEM image in figure

6-2D shows the recovered 2D highly ordered macroporous structure in the leaf region with the robust antireflection property. In addition, the highly ordered macroporous region recovered by contact pressure of the leaf patterned film exhibits optical characteristics as indicated in figure 6-4A. The average transmission spectra of the patterned area are 98.3% in the range of 425nm to 800nm, with a maximum of 99% at

475nm. In contrast, the cross-section SEM image in figure 6-2C shows the deformed macroporous structure with highly wrinkled and distorted arrangement in the rest area of the leaf region that exhibits a strong glare effect. The transmission substantially reduces to the 94.5%, which can be confirmed by the optical spectra in figure 6-4A. Both of them are dramatically higher (3%-8%) than the uncoated substrate that only has 91.0%. To characterize the reversibility, durability and reproducibility of transmission of the SMP, we used the normal incident angle spectrometer to cyclically monitor the transmission spectra when the SMP membrane dried out of water and pressure-induced recovery more than 100 times without any obvious degradation. The distinguishable zigzag transmission amplitudes that prove to be convincing evident are shown in figure 6-4B which show the average of absolute value (98.4%) of transmission of pressure-induced

SMP is apparently higher than the one (94.7%) dried out of water.

Atomic force microscopy (AFM) images of our SMP reveal the micropores that are reversibly manipulated by the external media during the cycling procedure.

Compared to the AFM images and the depth of profile in figure 6-3A, C and figure 6-3B,

D. It clearly shows that the pressure-triggered macroporous membrane has much

127

smoother superficial than the water dried film. Table 6-1, collectives all the surface roughness of SMP, shows broad variation from pressure-triggered SMP (RMS 3.3±0.9) to the SMP dried out of water(RMS 12.4±2.4).

In the process of pressure-triggered macroporous recovery, the entropy elasticity is the animated root affect the SM effect as described by our previous work. The cross- linked polymer chains were deployed at a stress-free arrangement which was highly energetically favorable. During the procedure of capillary pressure-triggered unusual cold programming, the considerable excess stresses are stored into the deformed polymer chains. The predetermined polymer networks have an extensive tendency to snap back to their stress-free state which has the maximum entropy as the stable state.

Thus, the extra van der Walls interaction between our stencil and SMP generate a small adhesive pull-off force, which can guide the spring-squeezed polymer chains to recover rapidly back to the stress-free state as proved by our previous nano-indentation test. In addition, based on the mechanism of pressure-triggering, we developed a variety of vivid pattern wisely employed by antireflection technology as indicated in figure 6-9 A,

Reconfigurable Antireflection 2D Photonic Crystals Enabled by Ethanol- Responsive Shape Memory Polymer

Against our expectations, the deformed tunable antireflection SMP can also recover by ethanol liquid. In this experiment, we chose the same sample as pressure- triggered SMP. The leaf pattern was disappeared when the membrane was dried out of water. Then, we used a Q-tip which was soaked up with ethanol to spread the half membrane. The photograph compares the area which is spread by ethanol standing in a vivid contrast to the area without spreading as shown in Figure 6-5A. The cross-

128

section SEM images in figure 6-5B show the region which dried out of water on the right side of figure 6-5A. Simultaneously, the SEM images in figure 6-5C of ethanol-treated

SMP which obtained on the left side of figure 6-5A shows much more ordered arrangement than the sample dried out of water. In the AFM images and surface profiler, the ethanol-responsive area (figure 6-6A) shows much smoother surface than the sample out of water (figure 6-6B). This result also confirmed by the surface roughness analysis in table 6-1. The surface roughness (RMS) collected by AFM shows the sample dried out of water (12.4±2.4nm) has 2 times larger than the sample recovered by ethanol (6.1±0.8nm) and 4 times larger than the sample triggered by pressure(3.3±0.9nm).

According to this promising approach, we found another external stimulus to design the broadband tunable antireflection property by a simple organic aqueous treatment, such as ethanol, acetone, isopropyl alcohol. Comparing the transmission of the porous film dried out of water and ethanol, the transmission spectra exhibit tremendous higher value (99.5%) treated with ethanol than water (94.5%) as dedicated in figure7a. The durable reversibility by drying out of water (94.5%) and ethanol (99.5%) according to priority are described by the amplitude of transmission measurement as indicated in figure 6-7B. We speculate the mechanism of ethanol-responsive AR film to the entropy-elasticity. The surface textures tend to redefine when the surface suffered by different external stimuli. The cumulative-entropy govern the surface porosity reshape and regulate the deformed microvoids recovered to the pop-up state. To clarify the postulation, we compare the surface tension of ethanol (22.39 m Nm-1 at 20ºC) and water (72.75 m Nm-1 at 20ºC).

129

Conclusion

In summary, our work presents a smart porous coating on polymer film. By integrating the fast-assembling 2D photonic crystals and shape memory polymers, the unique coating allows us to predictively manipulate the local macrospores collapse and reopen in terms of conspicuous variation of antireflection property. This variation enables a portable approach to accurate fine-turn the transmission efficiency, ranging from 94.5% to 99.5%. Moreover, the convenient unusual “cold” programming and unprecedented recovery condition achieved either by small contact pressure or ethanol spreading inspire us to further develop a smart optical device, such as artificial moth eyes, smart windows, and smart digital screen. On the other hand, implementation of versatility of surface structure is another significant approach for drug delivery, lock or unlock micro particles, tunable hydrophobic surface coating.

130

Figure 6-1. Schematic drawing the shape memory effect at nanoporous SMP film.

131

Figure 6-2. Arbitrary patterns printed on nanoporous SMP membrane. A).Photograoh of cubic pattern printed on water-dried nanoporous SMP film. B) photograph of leaf pattern printed on water-dried nanoporous SMP film. C) Top view SEM image of water-dried area in A). D) Top view SEM image of printed area in A). Photo courtesy of author.

132

Figure 6-3. Various morphology between water dried area and contact printed area. A).3-D AFM image of water dried sample of 100 nm nanoporous SMP film. B)3-D AFM image of printed sample of 100 nm nanoporous SMP film. C) Height profile scanned across the line for water dried sample in A). D) Height

profile scanned across the line for printed sample in B).

133

Figure 6-4. Normal-incidence optical transmission spectra obtained through a nanoporous SMP film of 100nm nanopores. A) Comparing the transmission spectra of pure polymer film, water dried nanoporous SMP film and printed nanoporous SMP film. B) Normal-incidence optical transmission spectra of 100nm nanopores SMP film dried out of water and printed by PDMS for 5 times. C) Transmission amplitudes of the spectra in B) at 500nm wavelength.

134

Figure 6-5. Nanoporous film rinsed by water and ethanol. A). Photograph of nanoporous film rinsed by water and ethanol. B) Top-view of SEM image water-dried area on the right side in A). C) Top view SEM image of ethanol-dried area on the left side in A). Photo courtesy of author.

135

Figure 6-6. Various morphology between water dried area and ethanol area. A) 3-D AFM image of water-dried sample of 100 nm nanoporous SMP film. B )3-D AFM image of ethanol-dried sample of 100 nm nanoporous SMP film. C) Height profile scanned across the line for water-dried sample in A. D) Height

profile scanned across the line for ethanol-dried sample in B).

136

Figure 6-7. Normal-incidence optical transmission spectra obtained through a nanoporous SMP film of 100nm nanopores. A) Normal-incidence optical transmission spectra of 100nm nanopores SMP film dried out of water and ethanol for 5 times. B) Transmission amplitudes of the spectra in B) at 500nm wavelength.

137

Figure 6-8. Templated fabrication of close packed 100nm microsphere on glass substrate. A) Photograph of close packed 100nm microsphere on glass substrate. B) Top view SEM image in A). Photo courtesy of author.

138

Figure 6-9. Photograph of patterned structures on nanoporous SMP film. A) Patterned chess pad by directly printing on water-dried nanoporous SMP film. B) Patterned butterfly by directly printing on water-dried nanoporous SMP film. Photo courtesy of author.

139

Table 6-1. Linear profile roughness.

AA roughness, Sa (nm) RMS roughness, Sq (nm) PDMS 2.7 ± 0.8 3.3 ± 0.9 Ethanol 4.9 ± 0.7 6.1 ± 0.8 Water 10.2 ± 2.2 12.4 ± 2.4

140

CHAPTER 7 CONCLUSIONS AND FUTURE WORK

I summary, by Integrating the concept of self-assembly photonic crystals with shape memory polymers, we have developed a series of new type of stimuli-responsive

SMP that is remarkably different from current SMPs. They enable instantaneous shape memory response and all-room-temperature operations for the entire SM cycle. In addition, the entire shape memory effect can be tailored from macroscale to nanoscale.

Which renders striking color changes during the shape memory process. Moreover, the large capillary pressure caused by water evaporation from the templated macropores results in the unique ‘cold’ programming at room temperature. The instantaneous recovery of the temporarily deformed macropores to the permanent, 3D highly ordered photonic crystal structure can be triggered by applying a small contact pressure, small shear pressure at ambient conditions.

In addition to being pressure responsive, the disorder-to-order transition of the new SMPs can also be stimulated by drying the macroporous SMP membranes out of solvents with low surface tension, such as ethanol and toluene. Importantly, the easily perceived colour change from translucency to iridescence associated with the structural disorder-to-order transition enables a simple and quantitative optical technology for characterizing the intriguing SM effects at nanoscale. Simultaneously, the striking chromogenic effects induced by the recovery of the permanent 3D photonic crystal structure provide opportunities for a wide spectrum of applications ranging from reconfigurable photonic crystal devices to chromogenic pressure and chemical sensors to novel biometric and anti-counterfeiting materials.

141

LIST OF REFERENCES

1 Lendlein, A. & Kelch, S. Shape-memory polymers. Angew. Chem.-Int. Edit. 41, 2034-2057, doi:10.1002/1521-3773(20020617)41:12<2034::aid- anie2034>3.0.co;2-m (2002).

2 Liu, Y. P., Gall, K., Dunn, M. L. & McCluskey, P. Thermomechanics of shape memory polymer nanocomposites. Mech Mater 36, 929-940 (2004).

3 Madbouly, S. A. & Lendlein, A. Shape-Memory Polymer Composites. Adv Polym Sci 226, 41-95 (2010).

4 Lendlein, A. & Kelch, S. Shape-memory polymers as stimuli-sensitive implant materials. Clin Hemorheol Micro 32, 105-116 (2005).

5 Lendlein, A. & Langer, R. Biodegradable, elastic shape-memory polymers for potential biomedical applications. Science 296, 1673-1676, doi:DOI 10.1126/science.1066102 (2002).

6 Shape-memory polymer wire treats blood clots in stroke. Adv Mater Process 164, 80-80 (2006).

7 Hu, Y. Shape-memory polymer system developed for medical applications. Mrs Bull 27, 488-489 (2002).

8 El Feninat, F., Laroche, G., Fiset, M. & Mantovani, D. Shape memory materials for biomedical applications. Adv Eng Mater 4, 91-104 (2002).

9 Behl, M., Razzaq, M. Y. & Lendlein, A. Multifunctional Shape-Memory Polymers. Adv Mater 22, 3388-3410 (2010).

10 Huang, W. M., Yang, B., Zhao, Y. & Ding, Z. Thermo-moisture responsive polyurethane shape-memory polymer and composites: a review. J Mater Chem 20, 3367-3381 (2010).

11 Filion, T. M., Xu, J. W., Prasad, M. L. & Song, J. In vivo tissue responses to thermal-responsive shape memory polymer nanocomposites. Biomaterials 32, 985-991 (2011).

12 Behl, M. & Lendlein, A. Shape-memory polymers. Mater Today 10, 20-28 (2007). 13 Behl, M., Zotzmann, J. & Lendlein, A. Shape-Memory Polymers and Shape- Changing Polymers. Adv Polym Sci 226, 1-40 (2010).

14 Liu, C., Qin, H. & Mather, P. T. Review of progress in shape-memory polymers. J Mater Chem 17, 1543-1558 (2007).

142

15 Ionov, L. et al. Fast and spatially resolved environmental probing using stimuli- responsive polymer layers and fluorescent nanocrystals. Adv Mater 18, 1453-+ (2006).

16 Le, D. M., Tycon, M. A., Fecko, C. J. & Ashby, V. S. Near-Infrared Activation of Semi-Crystalline Shape Memory Polymer Nanocomposites. J Appl Polym Sci 130, 4551-4557 (2013).

17 Lee, K. M., Koerner, H., Vaia, R. A., Bunning, T. J. & White, T. J. Light-activated shape memory of glassy, azobenzene liquid crystalline polymer networks. Soft Matter 7, 4318-4324 (2011).

18 Lendlein, A., Jiang, H. Y., Junger, O. & Langer, R. Light-induced shape-memory polymers. Nature 434, 879-882 (2005).

19 Li, M. H. & Keller, P. Stimuli-responsive polymer vesicles. Soft Matter 5, 927-937 (2009).

20 Quitmann, D., Gushterov, N., Sadowski, G., Katzenberg, F. & Tiller, J. C. Solvent-Sensitive Reversible Stress-Response of Shape Memory Natural Rubber. Acs Appl Mater Inter 5, 3504-3507 (2013).

21 Lv, H. B., Leng, J. S., Liu, Y. J. & Du, S. Y. Shape-memory polymer in response to solution. Adv Eng Mater 10, 592-595 (2008).

22 Liu, Y. J., Lv, H. B., Lan, X., Leng, J. S. & Du, S. Y. Review of electro-active shape-memory polymer composite. Compos Sci Technol 69, 2064-2068 (2009).

23 Lu, H. B. & Gou, J. H. Study on 3-D High Conductive Graphene Buckypaper for Electrical Actuation of Shape Memory Polymer. Nanosci Nanotech Let 4, 1155- 1159 (2012).

24 Han, X. J. et al. pH-Induced Shape-Memory Polymers. Macromol Rapid Comm 33, 1055-1060 (2012).

25 Fu, C. C. et al. Tunable Nanowrinkles on Shape Memory Polymer Sheets. Adv Mater 21, 4472-+ (2009).

26 Conti, S., Lenz, M. & Rumpf, M. Macroscopic behaviour of magnetic shape- memory polycrystals and polymer composites. Mat Sci Eng a-Struct 481, 351- 355 (2008).

27 Cui, J., Kratz, K., Heuchel, M., Hiebl, B. & Lendlein, A. Mechanically active scaffolds from radio-opaque shape-memory polymer-based composites. Polym Advan Technol 22, 180-189 (2011).

143

28 Xie, T. Recent advances in polymer shape memory. Polymer 52, 4985-5000, doi:10.1016/j.polymer.2011.08.003 (2011).

29 DiOrio, A. M., Luo, X. F., Lee, K. M. & Mather, P. T. A functionally graded shape memory polymer. Soft Matter 7, 68-74 (2011).

30 Xie, T. Shape memory polymer: Fundamentals and applications. Abstr Pap Am Chem S 246 (2013).

31 Xie, T. Tunable polymer multi-shape memory effect. Nature 464, 267-270, doi:10.1038/nature08863 (2010).

32 Wu, Y. P. et al. Light-Induced Shape Memory Polymer Materials. Prog Chem 24, 2004-2010 (2012).

33 Ma, M. M., Guo, L., Anderson, D. G. & Langer, R. Bio-Inspired Polymer Composite Actuator and Generator Driven by Water Gradients. Science 339, 186-189 (2013).

34 Li, J. J., An, Y. H., Huang, R., Jiang, H. Q. & Xie, T. Unique Aspects of a Shape Memory Polymer As the Substrate for Surface Wrinkling. Acs Appl Mater Inter 4, 598-603, doi:10.1021/am201727a (2012).

35 Haines, C. S. et al. Artificial Muscles from Fishing Line and Sewing Thread. Science 343, 868-872 (2014).

36 Davies, D. J. D. et al. A Printable Optical Time-Temperature Integrator Based on Shape Memory in a Chiral Nematic Polymer Network. Adv Funct Mater 23, 2723- 2727 (2013).

37 Tokudome, Y., Suzuki, K., Kitanaga, T. & Takahashi, M. Hierarchical Nested Wrinkles on Silica-Polymer Hybrid Films: Stimuli-Responsive Micro Periodic Surface Architectures. Sci Rep-Uk 2 (2012).

38 Sun, L. et al. Stimulus-responsive shape memory materials: A review. Mater Design 33, 577-640 (2012).

39 Pretsch, T., Ecker, M., Schildhauer, M. & Maskos, M. Switchable information carriers based on shape memory polymer. J Mater Chem 22, 7757-7766 (2012). 40 Meng, Q. H. & Hu, J. L. A review of shape memory polymer composites and blends. Compos Part a-Appl S 40, 1661-1672 (2009).

41 Mendez, J. et al. Bioinspired Mechanically Adaptive Polymer Nanocomposites with Water-Activated Shape-Memory Effect. Macromolecules 44, 6827-6835 (2011).

144

42 Meng, H. & Li, G. Q. A review of stimuli-responsive shape memory polymer composites. Polymer 54, 2199-2221 (2013).

43 Li, W. B. et al. Tuning surface micropattern features using a shape memory functional polymer. Rsc Adv 3, 9865-9874 (2013).

44 Eisenhaure, J. D., Xie, T., Varghese, S. & Kim, S. Microstructured Shape Memory Polymer Surfaces with Reversible Dry Adhesion. Acs Appl Mater Inter 5, 7714-7717, doi:10.1021/am402479f (2013).

45 Alarcon, C. D. H., Farhan, T., Osborne, V. L., Huck, W. T. S. & Alexander, C. Bioadhesion at micro-patterned stimuli-responsive polymer brushes. J Mater Chem 15, 2089-2094 (2005).

46 Forchel, A. Photonic crystal: Switching light with light. Nat Mater 2, 13-14 (2003). 47 Johnson, S. G., Mekis, A., Fan, S. H. & Joannopoulos, J. D. Molding the flow of light. Comput Sci Eng 3, 38-47 (2001).

48 Arsenault, A. C. et al. From colour fingerprinting to the control of photoluminescence in elastic photonic crystals. Nat Mater 5, 179-184 (2006).

49 Lee, J. H. et al. 25th Anniversary Article: Ordered Polymer Structures for the Engineering of Photons and Phonons. Adv Mater 26, 532-568, doi:10.1002/adma.201303456 (2014).

50 Thomas, E. L. Periodic bicontinuous polymer nanostructures. Abstr Pap Am Chem S 245 (2013).

51 Ledermann, A. et al. Three-dimensional silicon inverse photonic quasicrystals for infrared wavelengths. Nat Mater 5, 942-945 (2006).

52 Choi, S. Y., Mamak, M., von Freymann, G., Chopra, N. & Ozin, G. A. Mesoporous Bragg stack color tunable sensors. Nano Lett 6, 2456-2461 (2006). 53 Moon, J. H. & Yang, S. Chemical Aspects of Three-Dimensional Photonic Crystals. Chem Rev 110, 547-574 (2010).

54 Thomas, E. L. et al. Polymer based photonic band gap structures: 3D interference lithography, self assembly and 2 photon properties. Abstr Pap Am Chem S 229, U717-U717 (2005).

55 Bockstaller, M. R. & Thomas, E. L. Optical properties of polymer-based photonic nanocomposite materials. J Phys Chem B 107, 10017-10024, doi:10.1021/jp035286j (2003).

145

56 Fang, Y. et al. Scalable bottom-up fabrication of colloidal photonic crystals and periodic plasmonic nanostructures. J Mater Chem C 1, 6031-6047, doi:10.1039/c3tc30740a (2013).

57 Yang, H. T., Gozubenli, N., Fang, Y. & Jiang, P. Generalized Fabrication of Monolayer Nonclose-Packed Colloidal Crystals with Tunable Lattice Spacing. Langmuir 29, 7674-7681, doi:10.1021/la4011554 (2013).

58 Song, Y. M. et al. Digital cameras with designs inspired by the arthropod eye. Nature 497, 95-99, doi:10.1038/nature12083 (2013).

59 Askar, K. et al. Self-assembled self-cleaning broadband anti-reflection coatings. Colloid Surface A 439, 84-100, doi:10.1016/j.colsurfa.2013.03.004 (2013).

60 Potyrailo, R. A. et al. Towards outperforming conventional sensor arrays with fabricated individual photonic vapour sensors inspired by Morpho butterflies. Nat Commun 6, doi:ARTN 795910.1038/ncomms8959 (2015).

61 Yu, C. J. et al. Adaptive optoelectronic camouflage systems with designs inspired by cephalopod skins. P Natl Acad Sci USA 111, 12998-13003, doi:10.1073/pnas.1410494111 (2014).

62 Mahdavi, A. et al. A biodegradable and biocompatible gecko-inspired tissue adhesive. P Natl Acad Sci USA 105, 2307-2312, doi:10.1073/pnas.0712117105 (2008).

63 Mohammed, J. S. & Murphy, W. L. Bioinspired Design of Dynamic Materials. Adv Mater 21, 2361-2374, doi:10.1002/adma.200803785 (2009).

64 Ortiz, C. & Boyce, M. C. Materials science - Bioinspired structural materials. Science 319, 1053-1054, doi:10.1126/science.1154295 (2008).

65 Hellmich, C. & Katti, D. Mechanics of Biological and Bioinspired Materials and Structures. J Eng Mech-Asce 135, 365-366, doi:10.1061/(Asce)0733- 9399(2009)135:5(365) (2009).

66 Demirel, M. C., Cetinkaya, M., Pena-Francesch, A. & Jung, H. H. Recent Advances in Nanoscale Bioinspired Materials. Macromol Biosci 15, 300-311, doi:10.1002/mabi.201400324 (2015).

67 Chen, P. Y., McKittrick, J. & Meyers, M. A. Biological materials: Functional adaptations and bioinspired designs. Prog Mater Sci 57, 1492-1704, doi:10.1016/j.pmatsci.2012.03.001 (2012).

146

68 Yang, H. T., Dou, X., Fang, Y. & Jiang, P. Self-assembled biomimetic superhydrophobic hierarchical arrays. J Colloid Interf Sci 405, 51-57, doi:10.1016/j.jcis.2013.05.040 (2013).

69 Teyssier, J., Saenko, S. V., van der Marel, D. & Milinkovitch, M. C. Photonic crystals cause active colour change in chameleons. Nat Commun 6, doi:ARTN 636810.1038/ncomms7368 (2015).

70 Hearon, K. et al. Porous Shape-Memory Polymers. Polym Rev 53, 41-75, doi:10.1080/15583724.2012.751399 (2013).

71 Hu, J. L., Zhu, Y., Huang, H. H. & Lu, J. Recent advances in shape-memory polymers: Structure, mechanism, functionality, modeling and applications. Prog Polym Sci 37, 1720-1763, doi:10.1016/j.progpolymsci.2012.06.001 (2012).

72 Li, J. J. & Xie, T. Significant Impact of Thermo-Mechanical Conditions on Polymer Triple-Shape Memory Effect. Macromolecules 44, 175-180, doi:10.1021/ma102279y (2011).

73 Ratna, D. & Karger-Kocsis, J. Recent advances in shape memory polymers and composites: a review. J Mater Sci 43, 254-269, doi:10.1007/s10853-007-2176-7 (2008).

74 Kim, B. K. Shape memory polymers and their future developments. Express Polym Lett 2, 614-614, doi:10.3144/expresspolymlett.2008.73 (2008).

75 Dietsch, B. & Tong, T. A review - Features and benefits of shape memory polymers (SMPs). J Adv Mater-Covina 39, 3-12 (2007).

76 Jiang, P., Bertone, J. F., Hwang, K. S. & Colvin, V. L. Single crystal colloidal multilayers of controlled thickness. Abstr Pap Am Chem S 218, U431-U431 (1999).

77 Menzies, A. C. & Skinner, J. The Growing of Crystals - by the Methods of Kyropoulos and of Stober. Discuss Faraday Soc, 306-312, doi:DOI 10.1039/df9490500306 (1949).

78 Halford, B. Photonic Crystals. Chem Eng News 93, 5-5 (2015). 79 Joannopoulos, J. D., Villeneuve, P. R. & Fan, S. H. Photonic crystals. Solid State Commun 102, 165-173, doi:Doi 10.1016/S0038-1098(96)00716-8 (1997).

80 Pradhan, R. D., Bloodgood, J. A. & Watson, G. H. Photonic band structure of bcc colloidal crystals. Phys Rev B 55, 9503-9507, doi:DOI 10.1103/PhysRevB.55.9503 (1997).

147

81 Yablonovitch, E. Photonic Crystals. J Mod Optic 41, 173-194, doi:Doi 10.1080/09500349414550261 (1994).

82 Oliver, W. C. & Pharr, G. M. Measurement of hardness and elastic modulus by instrumented indentation: Advances in understanding and refinements to methodology. J. Mater. Res. 19, 3-20, doi:10.1557/jmr.2004.0002 (2004).

83 Ahn, S. K., Kasi, R. M., Kim, S. C., Sharma, N. & Zhou, Y. X. Stimuli-responsive polymer gels. Soft Matter 4, 1151-1157 (2008).

84 Du, H. Y. & Zhang, J. H. Solvent induced shape recovery of shape memory polymer based on chemically cross-linked poly(vinyl alcohol). Soft Matter 6, 3370-3376 (2010).

85 Fang, Y. et al. Reconfigurable photonic crystals enabled by pressure-responsive shape-memory polymers. Nat Commun 6, doi:ARTN 741610.1038/ncomms8416 (2015).

86 Gu, X. Z. & Mather, P. T. Water-triggered shape memory of multiblock thermoplastic polyurethanes (TPUs). Rsc Adv 3, 15783-15791 (2013).

87 Iqbal, D. & Samiullah, M. H. Photo-Responsive Shape-Memory and Shape- Changing Liquid-Crystal Polymer Networks. Materials 6, 116-142 (2013).

88 Razzaq, M. Y., Anhalt, M., Frormann, L. & Weidenfeller, B. Thermal, electrical and magnetic studies of magnetite filled polyurethane shape memory polymers. Mat Sci Eng a-Struct 444, 227-235, doi:10.1016/j.msea.2006.08.083 (2007).

89 Xu, H., Norsten, T. B., Uzun, O., Jeoung, E. & Rotello, V. M. Stimuli responsive surfaces through recognition-mediated polymer modification. Chem Commun, 5157-5159 (2005).

90 Howarter, J. A. & Youngblood, J. P. Self-cleaning and anti-fog surfaces via stimuli-responsive polymer brushes. Adv Mater 19, 3838-+ (2007).

91 Sokolowski, W., Metcalfe, A., Hayashi, S., Yahia, L. & Raymond, J. Medical applications of shape memory polymers. Biomed Mater 2, S23-S27, doi:Doi 10.1088/1748-6041/2/1/S04 (2007).

92 Berger, S., Synytska, A., Ionov, L., Eichhorn, K. J. & Stamm, M. Stimuli- Responsive Bicomponent Polymer Janus Particles by "Grafting from"/"Grafting to" Approaches. Macromolecules 41, 9669-9676 (2008).

93 Stuart, M. A. C. et al. Emerging applications of stimuli-responsive polymer materials. Nat Mater 9, 101-113 (2010).

148

94 Wischke, C., Neffe, A. T. & Lendlein, A. Controlled Drug Release from Biodegradable Shape-Memory Polymers. Adv Polym Sci 226, 177-205 (2010).

95 Behl, M., Kratz, K., Zotzmann, J., Nochel, U. & Lendlein, A. Reversible Bidirectional Shape-Memory Polymers. Adv Mater 25, 4466-4469, doi:10.1002/adma.201300880 (2013).

96 Safranski, D. L., Smith, K. E. & Gall, K. Mechanical Requirements of Shape- Memory Polymers in Biomedical Devices. Polym Rev 53, 76-91, doi:10.1080/15583724.2012.752385 (2013).

97 Jiang, P., Hwang, K. S., Mittleman, D. M., Bertone, J. F. & Colvin, V. L. Template-directed preparation of macroporous polymers with oriented and crystalline arrays of voids. J Am Chem Soc 121, 11630-11637 (1999).

98 Nair, R. V. & Jagatap, B. N. Bragg wave coupling in self-assembled opal photonic crystals. Phys Rev A 85, doi:ARTN 01382910.1103/PhysRevA.85.013829 (2012).

99 Song, B. S., Noda, S. & Asano, T. Photonic devices based on in-plane hetero photonic crystals. Science 300, 1537-1537, doi:DOI 10.1126/science.1083066 (2003).

100 Tetreault, N., Miguez, H., Yang, S. M., Kitaev, V. & Ozin, G. A. Refractive index patterns in silicon inverted colloidal photonic crystals. Adv Mater 15, 1167-+, doi:10.1002/adma.200304429 (2003).

101 Ogawa, S. P., Imada, M., Yoshimoto, S., Okano, M. & Noda, S. Control of light emission by 3D photonic crystals. Science 305, 227-229, doi:10.1126/science.1097968 (2004).

102 Lopez, C. Materials aspects of photonic crystals. Adv Mater 15, 1679-1704, doi:10.1002/adma.200300386 (2003).

103 Dong, W. T., Bongard, H., Tesche, B. & Marlow, F. Inverse opals with a skeleton structure: Photonic crystals with two complete bandgaps. Adv Mater 14, 1457-+, doi:Doi 10.1002/1521-4095(20021016)14:20<1457::Aid-Adma1457>3.0.Co;2-P (2002).

104 Yablonovitch, E. Photonic crystals: Semiconductors of light. Sci Am 285, 46-+ (2001).

105 Chen, Z. H., Tan, Q. L., Lao, J., Liang, Y. & Huang, X. G. Reconfigurable and tunable flat graphene photonic crystal circuits. Nanoscale 7, 10912-10917, doi:10.1039/c5nr01343g (2015).

149

106 Gaughan, R. Researchers create tunable photonic bandgap crystal. Photon Spectra 34, 26-26 (2000).

107 Gur, D. et al. The Mechanism of Color Change in the Neon Tetra Fish: a Light- Induced Tunable Photonic Crystal Array. Angew. Chem.-Int. Edit. 54, 12426- 12430, doi:10.1002/anie.201502268 (2015).

108 Han, M. G. et al. Full Color Tunable Photonic Crystal from Crystalline Colloidal Arrays with an Engineered Photonic Stop-Band. Adv Mater 24, 6438-6444, doi:10.1002/adma.201203211 (2012).

109 Kim, H. et al. Structural colour printing using a magnetically tunable and lithographically fixable photonic crystal. Nat Photonics 3, 534-540, doi:10.1038/Nphoton.2009.141 (2009).

110 Liu, C. Y. & Chen, L. W. Tunable band gap in a photonic crystal modulated by a nematic liquid crystal. Phys Rev B 72, doi:ARTN 04513310.1103/PhysRevB.72.045133 (2005).

111 O'Brien, D. et al. Tunable optical delay using photonic crystal heterostructure nanocavities. Phys Rev B 76, doi:ARTN 11511010.1103/PhysRevB.76.115110 (2007).

112 Pustai, D. M., Sharkawy, A., Shi, S. Y. & Prather, D. W. Tunable photonic crystal microcavities. Appl Optics 41, 5574-5579, doi:Doi 10.1364/Ao.41.005574 (2002).

113 Shimamoto, N. et al. Nanopattern Fabrication of Gold on Hydrogels and Application to Tunable Photonic Crystal. Adv Mater 24, 5243-5248, doi:10.1002/adma.201201522 (2012).

114 Fang, Y. et al. Optically Bistable Macroporous Photonic Crystals Enabled by Thermoresponsive Shape Memory Polymers. Adv Opt Mater 3, 1509-1516, doi:10.1002/adom.201500277 (2015).

115 Kokkinis, D., Schaffner, M. & Studart, A. R. Multimaterial magnetically assisted 3D printing of composite materials. Nat Commun 6, doi:ARTN 864310.1038/ncomms9643 (2015).

116 Kim, J. H. et al. 3D Printing of Reduced Graphene Oxide Nanowires. Adv Mater 27, 157-161, doi:10.1002/adma.201404380 (2015).

117 Lao, Z. X. et al. Capillary Force Driven Self-Assembly of Anisotropic Hierarchical Structures Prepared by Femtosecond Laser 3D Printing and Their Applications in Crystallizing Microparticles. Acs Nano 9, 12060-12069, doi:10.1021/acsnano.5b04914 (2015).

150

118 Gratson, G. M. et al. Direct-write assembly of three-dimensional photonic crystals: Conversion of polymer scaffolds to silicon hollow-woodpile structures. Adv Mater 18, 461-+, doi:10.1002/adma.200501447 (2006).

119 Eckel, Z. C. et al. 3D PRINTING Additive manufacturing of polymer-derived ceramics. Science 351, 58-62, doi:10.1126/science.aad2688 (2016).

120 Hong, S. M. et al. 3D Printing of Highly Stretchable and Tough Hydrogels into Complex, Cellularized Structures. Adv Mater 27, 4035-4040, doi:10.1002/adma.201501099 (2015).

121 Takahashi, S. et al. Direct creation of three-dimensional photonic crystals by a top-down approach. Nat Mater 8, 721-725, doi:10.1038/Nmat2507 (2009).

122 Wu, J. T. et al. Multi-shape active composites by 3D printing of digital shape memory polymers. Sci Rep-Uk 6, doi:ARTN 2422410.1038/srep24224 (2016).

123 Wong, S. et al. Direct laser writing of three-dimensional photonic crystals with complete a photonic bandgap in chalcogenide glasses. Adv Mater 18, 265-+, doi:10.1002/adma.200501973 (2006).

124 Zarek, M. et al. 3D Printing of Shape Memory Polymers for Flexible Electronic Devices. Adv Mater 28, 4449-+, doi:10.1002/adma.201503132 (2016).

125 Zhang, Q. Q. et al. 3D Printing of Graphene Aerogels. Small 12, 1702-1708, doi:10.1002/smll.201503524 (2016).

126 Raney, J. R. & Lewis, J. A. Printing mesoscale architectures. Mrs Bull 40, 943- 950, doi:10.1557/mrs.2015.235 (2015).

127 Sun, L. et al. Direct-Write Assembly of 3D Silk/Hydroxyapatite Scaffolds for Bone Co-Cultures. Adv Healthc Mater 1, 729-735, doi:10.1002/adhm.201200057 (2012).

128 Horiuchi, N. PHOTONIC CRYSTALS Bridging the visible. Nat Photonics 7, 767- 767, doi:10.1038/nphoton.2013.255 (2013).

129 Kopnov, F., Lirtsman, V. & Davidov, D. Self-assembled colloidal photonic crystals. Synthetic Met 137, 993-995, doi:10.1016/S0379-6779(02)00839-1 (2003).

130 Yablonovitch, E. Photonic crystals - Towards rational material design. Nat Mater 2, 648-649, doi:10.1038/nmat987 (2003).

131 Hochrainer, D. & Stober, W. Stober-Rotor with Recirculation of Particle-Free Air. Am Ind Hyg Assoc J 39, 754-757, doi:Doi 10.1080/0002889778507846 (1978).

151

132 Maldovan, M. & Thomas, E. L. Diamond-structured photonic crystals. Nat Mater 3, 593-600, doi:10.1038/nmat1201 (2004).

133 Meade, S. O., Yoon, M. S., Ahn, K. H. & Sailor, M. J. Porous silicon photonic crystals as encoded microcarriers. Adv Mater 16, 1811-+, doi:10.1002/adma.200400713 (2004).

134 O'Regan, B. J., Wang, Y. & Krauss, T. F. Silicon photonic crystal thermal emitter at near-infrared wavelengths. Sci Rep-Uk 5, doi:ARTN 1341510.1038/srep13415 (2015).

135 Paderni, K. et al. Shape-memory polymer networks from sol-gel cross-linked alkoxysilane-terminated poly(epsilon-caprolactone). J Mater Sci 47, 4354-4362 (2012).

136 Pan, G. H., Huang, W. M., Ng, Z. C., Liu, N. & Phee, S. J. The glass transition temperature of polyurethane shape memory polymer reinforced with treated/non- treated attapulgite (playgorskite) clay in dry and wet conditions. Smart Materials & Structures 17 (2008).

137 Baer, G. M. et al. Thermomechanical Properties, Collapse Pressure, and Expansion of Shape Memory Polymer Neurovascular Stent Prototypes. J Biomed Mater Res B 90B, 421-429 (2009).

138 Basit, A., L'Hostis, G. & Durand, B. Multi-shape memory effect in shape memory polymer composites. Mater Lett 74, 220-222 (2012).

139 Lee, S. K., Park, H. S., Yi, G. R., Moon, J. H. & Yang, S. M. Holographic Fabrication of Microstructures with Internal Nanopatterns Using Microprism Arrays. Angew. Chem.-Int. Edit. 48, 7000-7005, doi:10.1002/anie.200901166 (2009).

140 Lu, Y., Yin, Y. D. & Xia, Y. N. A self-assembly approach to the fabrication of patterned, two-dimensional arrays of microlenses of organic polymers. Adv Mater 13, 34-+, doi:Doi 10.1002/1521-4095(200101)13:1<34::Aid-Adma34>3.3.Co;2-T (2001).

141 Hoenders, B. J., Doosje, M. & Knoester, J. Scattering of light by photonic crystals. Acta Phys Pol A 101, 133-146 (2002).

142 Gur, D. et al. Light-Induced Color Change in the Sapphirinid Copepods: Tunable Photonic Crystals. Adv Funct Mater 26, 1393-1399, doi:10.1002/adfm.201504339 (2016).

143 Braun, P. V. Pattern definition in self-assembled photonic crystals. Abstr Pap Am Chem S 223, D20-D21 (2002).

152

144 Shkunov, M. N. et al. Tunable, gap-state lasing in switcbable directions for opal photonic crystals. Adv Funct Mater 12, 21-26, doi:Doi 10.1002/1616- 3028(20020101)12:1<21::Aid-Adfm21>3.0.Co;2-S (2002).

145 Fenzl, C., Hirsch, T. & Wolfbeis, O. S. Photonic Crystals for Chemical Sensing and Biosensing. Angew. Chem.-Int. Edit. 53, 3318-3335 (2014).

146 Ge, J. P. & Yin, Y. D. Responsive Photonic Crystals. Angew. Chem.-Int. Edit. 50, 1492-1522, doi:10.1002/anie.200907091 (2011).

147 Hidalgo, N., Calvo, M. E. & Miguez, H. Mesostructured Thin Films as Responsive Optical Coatings of Photonic Crystals. Small 5, 2309-2315, doi:10.1002/smll.200900411 (2009).

148 Hu, H. B., Chen, Q. W., Tang, J., Hu, X. Y. & Zhou, X. H. Photonic anti- counterfeiting using structural colors derived from magnetic-responsive photonic crystals with double photonic bandgap heterostructures. J Mater Chem 22, 11048-11053, doi:10.1039/c2jm30169e (2012).

149 Li, C. & Lotsch, B. V. Stimuli-responsive 2D polyelectrolyte photonic crystals for optically encoded pH sensing. Chem Commun 48, 6169-6171, doi:10.1039/c2cc31916k (2012).

150 Fang, Y. et al. Chromogenic Photonic Crystals Enabled by Novel Vapor- Responsive Shape-Memory Polymers. Adv Mater 27, 3696-3704, doi:10.1002/adma.201500835 (2015).

151 Wang, Z. H. et al. Colorful detection of organic solvents based on responsive organic/inorganic hybrid one-dimensional photonic crystals. J Mater Chem 21, 1264-1270 (2011).

152 Wang, Z. H. et al. Bioinspired Water-Vapor-Responsive Organic/Inorganic Hybrid One-Dimensional Photonic Crystals with Tunable Full-Color Stop Band. Adv Funct Mater 20, 3784-3790 (2010).

153 Wang, L. B., Wang, J. X. & Song, Y. L. Research Progress of Fast-Responsive Polymer Photonic Crystals. Acta Polym Sin, 1118-1127, doi:10.3724/Sp.J.1105.2012.12133 (2012).

154 Ge, J. P. & Yin, Y. D. Magnetically responsive colloidal photonic crystals. J Mater Chem 18, 5041-5045, doi:10.1039/b809958h (2008).

155 Diao, Y. Y., Liu, X. Y., Toh, G. W., Shi, L. & Zi, J. Multiple Structural Coloring of Silk-Fibroin Photonic Crystals and Humidity-Responsive Color Sensing. Adv Funct Mater 23, 5373-5380, doi:10.1002/adfm.201203672 (2013).

153

156 Fang, Y. et al. Direct Writing of Three-Dimensional Macroporous Photonic Crystals on Pressure-Responsive Shape Memory Polymers. Acs Appl Mater Inter 7, 23650-23659, doi:10.1021/acsami.5b07220 (2015).

157 Ding, T., Smoukov, S. K. & Baumberg, J. J. Stamping colloidal photonic crystals: a facile way towards complex pixel colour patterns for sensing and displays. Nanoscale 7, 1857-1863, doi:10.1039/c4nr05934d (2015).

158 Domeier, L., Nissen, A., Goods, S., Whinnery, L. & McElhanon, J. Thermomechanical Characterization of Thermoset Urethane Shape-Memory Polymer Foams. J Appl Polym Sci 115, 3217-3229 (2010).

159 Sun, J., Liu, Y. Y. & Leng, J. S. Mechanical properties of shape memory polymer composites enhanced by elastic fibers and their application in variable stiffness morphing skins. J Intel Mat Syst Str 26, 2020-2027, doi:10.1177/1045389x14546658 (2015).

160 Zheng, N., Fang, G. Q., Cao, Z. L., Zhao, Q. & Xie, T. High strain epoxy shape memory polymer. Polym Chem-Uk 6, 3046-3053, doi:10.1039/c5py00172b (2015).

161 Zhao, Q., Qi, H. J. & Xie, T. Recent progress in shape memory polymer: New behavior, enabling materials, and mechanistic understanding. Prog Polym Sci 49-50, 79-120, doi:10.1016/j.progpolymsci.2015.04.001 (2015).

162 Zhao, Y., Bai, S. Y., Dumont, D. & Galstian, T. V. Mechanically tunable diffraction gratings recorded on an azobenzene elastomer. Adv Mater 14, 512-+, doi:Doi 10.1002/1521-4095(20020404)14:7<512::Aid-Adma512>3.3.Co;2-3 (2002).

163 Huang, L. L. et al. Broadband Hybrid Holographic Multiplexing with Geometric Metasurfaces. Adv Mater 27, 6444-+, doi:10.1002/adma.201502541 (2015).

164 Wu, M. H. & Whitesides, G. M. Fabrication of two-dimensional arrays of microlenses and their applications in photolithography. J Micromech Microeng 12, 747-758, doi:Pii S0960-1317(02)32053-9Doi 10.1088/0960-1317/12/6/305 (2002).

165 Xu, D. W., Graugnard, E., King, J. S., Zhong, L. W. & Summers, C. J. Large- scale fabrication of ordered nanobowl arrays. Nano Lett 4, 2223-2226, doi:10.1021/nl048589d (2004).

166 Sur, U. K. Stimuli-responsive bio-inspired synthetic polymer nanocomposites. Curr Sci India 98, 1562-1563 (2010).

154

167 Tolley, M. T. et al. Self-folding origami: shape memory composites activated by uniform heating. Smart Mater Struct 23, doi:Artn 09400610.1088/0964- 1726/23/9/094006 (2014).

168 Ahn, H. et al. Nanoporous Block Copolymer Membranes for Ultrafiltration: A Simple Approach to Size Tunability. Acs Nano 8, 11745-11752, doi:10.1021/nn505234v (2014).

169 Cai, J. G. et al. Self-cleaning, broadband and quasi-omnidirectional antireflective structures based on mesocrystalline rutile TiO2 nanorod arrays. Energ Environ Sci 5, 7575-7581, doi:10.1039/c2ee03573a (2012).

170 Fan, P. X. et al. Broadband High-Performance Infrared Antireflection Nanowires Facilely Grown on Ultrafast Laser Structured Cu Surface. Nano Letters 15, 5988- 5994, doi:10.1021/acs.nanolett.5b02141 (2015).

171 Guldin, S. et al. Self-Cleaning Antireflective Optical Coatings. Nano Lett 13, 5329-5335, doi:10.1021/nl402832u (2013).

172 Jiang, H., Zhao, W. Z., Li, C. L. & Wang, Y. C. Polymer nanoparticle-based porous antireflective coating on flexible plastic substrate. Polymer 52, 778-785, doi:10.1016/j.polymer.2010.12.036 (2011).

173 Li, X., Gao, J. P., Xue, L. J. & Han, Y. C. Porous Polymer Films with Gradient- Refractive-Index Structure for Broadband and Omnidirectional Antireflection Coatings. Adv Funct Mater 20, 259-265, doi:10.1002/adfm.200901052 (2010).

174 Li, Y. F., Zhang, J. H. & Yang, B. Antireflective surfaces based on biomimetic nanopillared arrays. Nano Today 5, 117-127, doi:10.1016/j.nantod.2010.03.001 (2010).

175 Li, X. & Han, Y. C. Tunable wavelength antireflective film by non-solvent-induced phase separation of amphiphilic block copolymer micelle solution. J Mater Chem 21, 18024-18033, doi:10.1039/c1jm13704b (2011).

176 Mizoshita, N., Ishii, M., Kato, N. & Tanaka, H. Hierarchical Nanoporous Silica Films for Wear Resistant Antireflection Coatings. Acs Appl Mater Inter 7, 19424- 19430, doi:10.1021/acsami.5b05659 (2015).

177 Raut, H. K. et al. Multiscale Ommatidial Arrays with Broadband and Omnidirectional Antireflection and Antifogging Properties by Sacrificial Layer Mediated Nanoimprinting. Acs Nano 9, 1305-1314, doi:10.1021/nn5051272 (2015).

178 Shi, E. Z. et al. Colloidal Antireflection Coating Improves Graphene-Silicon Solar Cells. Nano Lett 13, 1776-1781, doi:10.1021/nl400353f (2013).

155

179 Spinelli, P., Verschuuren, M. A. & Polman, A. Broadband omnidirectional antireflection coating based on subwavelength surface Mie resonators. Nat Commun 3, doi:Artn 69210.1038/Ncomms1691 (2012).

180 Du, Y., Luna, L. E., Tan, W. S., Rubner, M. F. & Cohen, R. E. Hollow Silica Nanoparticles in UV-Visible Antireflection Coatings for Poly(methyl methacrylate) Substrates. Acs Nano 4, 4308-4316, doi:10.1021/nn101033y (2010).

181 Hiller, J., Mendelsohn, J. D. & Rubner, M. F. Reversibly erasable nanoporous anti-reflection coatings from polyelectrolyte multilayers. Nat Mater 1, 59-63, doi:10.1038/nmat719 (2002).

182 Kim, M. K., Yi, D. K. & Paik, U. Tunable, Flexible Antireflection Layer of ZnO Nanowires Embedded in PDMS. Langmuir 26, 7552-7554, doi:10.1021/la9043599 (2010).

183 Li, X., Yu, X. H. & Han, Y. C. Polymer thin films for antireflection coatings. J Mater Chem C 1, 2266-2285, doi:10.1039/c2tc00529h (2013).

184 Han, L. et al. Geometrically Controlled Mechanically Responsive Polyelectrolyte Tube Arrays. Adv Mater 23, 4667-+, doi:10.1002/adma.201102917 (2011).

185 Kumeria, T. et al. Photoswitchable Membranes Based on Peptide-Modified Nanoporous Anodic Alumina: Toward Smart Membranes for On-Demand Molecular Transport. Adv Mater 27, 3019-3024, doi:10.1002/adma.201500473 (2015).

156

BIOGRAPHICAL SKETCH

Yin Fang was born at Inner Mongolia in China. He graduated in August 2016 with his Ph.D degree. He has studied in University of Florida for five years. His major is chemical engineering. His research is related to smart nanomaterials, nanoscasle shape memory polymers, self-assemble photonic crystals, antireflection coatings, smart windows. He has published 6 first author papers and 6 co-author papers. He won the silver medal of graduated student award in materials research society in 2015.

157