Multi-Functional Stimuli-Responsive Materials

The first major project undertaken during this grant was the development of shape memory and healable polymers using disulfide bonds as a dynamic bonding motif. The disulfide bonds in the material can undergo dynamic exchange in response to a stimulus such as UV light or high temperatures (>150 °C) which would allow the material to flow and heal scratches or defects (Figure 1a). These materials are also semicrystalline crosslinked networks; when heated above the melting transition of the crystalline domains (~60 °C) the material would behavior as an elastomer and could be easily deformed. If the material was allowed to cool when in a deformed shape, it would recrystallize to lock in or “fix” that shape. Upon reheating above the melting transition of the crystalline domains the material recovers to its original shape (Figure 1b). The dynamic exchange of disulfides also allows the material to be reprocessed and have the shape permanently changed. This ability to “reprogram” the remembered shape is uncommon among shape memory materials and is an example of how the interplay of the healing mechanism and shape memory mechanism give this material properties unique properties. Another example of this interplay between the healing and shape memory mechanisms is through the use of shape memory assisted self-healing (SMASH) (Figure 1c). In this mechanism a material is previously deformed by heating above the melting transition, stretching the material, and then cooling it down in the stretched state, thereby fixing a tensile strain in the material. When the material is given a much wider scratch, the shape memory mechanism can be used to contract the material to its original length, which pushes the edges of the scratch closer together, effectively sealing the scratch. The scratch is subsequently healed using either UV light or higher temperatures (>150 °C). This seal-then-heal approach allows for the healing of much wider scratches or gouges in the material – including those in which material is removed – than could be achieved through the conventional healing mechanism. This work was published in ACS Macro Letters in 2013. In a related project these materials were demonstrated to function as reversible adhesives, wherein the softening of the material via melting of crystalline regions and exchange of disulfide bonds facilitated wetting onto a substrate. When the material is cooled and allowed to recrystallize a rigid adhesive bond is formed between the polymer and the substrate. It was demonstrated that the disulfide exchange allowed for superior adhesive properties over a nondynamic semicrystalline network, thereby proving that the dynamic network offers an advantage over traditional shape memory based adhesives. Furthermore, the shape memory effect of the adhesive allowed for the creation of shape memory joints where the relative configuration of two substrates joined with the adhesive could be altered and then returned to their original positions using the shape memory effect. A publication based on this work is in preparation. Multi-Responsive Actuating Materials from Liquid Crystalline Elastomers The second major project undertaken during the grant involved incorporating mesogenic ligands into a polymer network to create multiresponsive liquid crystalline elastomers (LCEs) which serve as actuating materials. The material incorporates a metal-binding ligand (2,6- bisbenzimidazolylpyridine (Bip)) which also can form thermotropic liquid crystals. When the ligand binds a metal ion it undergoes a conformational change that causes it to no longer be mesogenic. By incorporating these ligands into a polymer network, this transition from the ordered liquid crystalline state to the disordered isotropic state is translated to macroscopic actuation. Furthermore, the Bip mesogen is responsive to a number of stimuli: in addition to the aforementioned metal-binding response, the transition from liquid crystalline to isotropic can also be induced by heat (ca. 75 °C) or UV light (via a photo-thermal conversion). Thus, these materials serve as metallo-, thermo-, and photo-responsive actuators. In addition, the liquidcrystalline to isotropic transition can also be used as a shape memory trigger. If the material is deformed in the isotropic state and then allowed to transition back to liquid crystalline while deformed, it will lock in the deformation. When the material is returned to the isotropic state – either through heating or exposure to UV light or metal ions – it will recover to its original shape (Figure 2a). If the transition is achieved through metal binding, the metal can be removed using a competitive binder such as diethylene triamine. This effect is shown in figure 2a where the introduction of Fe2+ ions results in recovery to the original length as well as a purple color change in the material (caused by the Fe2+ binding with Bip). When the material is exposed to diethylene triamine the original color is restored, indicating that the Fe2+ has been removed. The metallo-responsive actuation of the material was demonstrated by placing a pre-strained sample of the material in 1:1 H2O:THF with a 10g mass attached. No actuation is observed when the material is placed in the solvent but once Fe(OTf)2 is added to make a 1mM solution the material noticeably contracts and lifts the mass (Figure 2b). Furthermore, the actuation stress of the material was measured using a tensile tester. For the metallo-stimulus (1mM Fe(OTf)2 in 1:1 H2O:THF) the actuation stress was 124 kPa and using a photo-stimulus (UV light, 320-390 nm, I=500 mW/cm2) the actuation stress was 212 kPa. The work on these materials was published in Macromolecules. One limitation of these materials is that they are polydomain liquid crystals. In polydomain liquid crystals the mesogens are grouped in multiple domains and orientational order is only preserved within a given domain and not across the entire material. Thus, upon undergoing the LC to isotropic transition the contractile forces of all the differently oriented domains largely cancel out resulting in minimal macroscopic actuation. This limitation can be overcome by producing monodomain liquid crystalline elastomers where all mesogens across the sample have the same orientation. In the previously mentioned materials, an external force was used to stretch the sample in the isotropic state. When the sample is cooled in the stretched state it will undergo the LC transition and produce a monodomain liquid crystal. The monodomain alignment however is lost when the sample is reheated to the isotropic state and then cooled without an external force. Another way to overcome this is by aligning the mesogens (via mechanical shear or magnetic alignment) prior to crosslinking, producing an intrinsically monodomain LCE. However, recently it has been shown that monodomain LCEs can be easily accessed in LCEs which contain dynamic bonds in the LCE backbone and/or crosslinks (Figure 3a). In this process the multi-domain material is stretched to align the mesogens while a stimulus is applied that induces the dynamic bonds in the material to undergo exchange reactions. The stimulus is then removed while the material is still in the stretched state allowing the material to “relax” in the monodomain morphology. We have begun producing LCE show that we can get repeated actuation with constant stress by heating and cooling around the LC transition Shape Memory Porous Materials Our final project was a cllaborative effort with Dr. Mary Ann Meador and the Aerogel group at NASA Glenn Research Center. The goal of this project is to create a shape memory aerogel. These materials would have the desired properties of aerogels (low density, low thermal conductivity) but also be able to display a shape memory effect similar to the dynamic materials we have worked with previously. Due to their large amount of free volume, aerogels can be compressed to very small sizes. Using the shape memory effect, they should be able to retain this small size until the application of a stimulus triggers the shape change back to their original size. This would be useful for deployable space structures, which can be transported at a size much smaller than their actual size. It would also be useful for expandable insulation, allowing for insulating layers to be placed between irregular surfaces. The materials were produced by synthesizing highly crosslinked thiol-ene networks. A dithiol (hexanedithiol, HDT) and a tetrathiol (pentaerythritoltetrakis(3-mercaptopropionate), PETMP) were reacted with a triallyl (1,3,5-Triallyl1,3,5-triazine-2,4,6-trione, TTT) using a redox activated thiol-ene reaction in acetonitrile or acetone followed by removal of the solvent using supercritical CO2 extraction to produce a highly porous material. The materials have a porosity of ca. 70-73% (for those produced in acetonitrile) and ca. 80-83% (for those produced in acetone). All the materials have a surface area of ca. 5-10 m2/g which is low for an aerogel material. The materials also have a microporous structure with pore sizes on the order of 1 µm (Figure 4a). The materials show good shape memory properties with the ability to fix a compressive strain of ca. 17% with >95% fixing and recovery (Figure 4b). A publication based on this work is in preparation. Also, a similar project using cellulose nanocrystals to produce nanocomposite aerogels stemmed from this work resulting in a patent application (Nguyen, B; Meador, M. A.; Rowan, S.; Cudjoe, E.; Michal, B. Polymer Cellulose Nanocrystal Composite Aerogels U.S. Patent Application No. 62,031,421, 2014).