Phototunable Cross-Linked Polysiloxanes - ACS Publications

Article pubs.acs.org/Macromolecules

Phototunable Cross-Linked Polysiloxanes Amanda S. Fawcett,† Timothy C. Hughes,‡ Laura Zepeda-Velazquez,† and Michael A. Brook*,† †

Department of Chemistry and Chemical Biology, McMaster University, 1280 Main Street West, Hamilton, ON, Canada L8S 4M1 CSIRO Manufacturing Flagship, Bayview Avenue, Clayton, VIC 3168, Australia

‡

Downloaded by UNIV OF SHEFFIELD on September 5, 2015 | http://pubs.acs.org Publication Date (Web): September 3, 2015 | doi: 10.1021/acs.macromol.5b01085

S Supporting Information *

ABSTRACT: Silicone elastomers are normally thermoset materials. While their inherent properties make them highly valuable, it would be of interest to develop stimuli-responsive silicones whose properties could be reversibly tuned at will. In the case of silicone polymers, a particularly interesting trigger is light, since silicone elastomers can readily be formulated to be transparent. We describe the utilization of coumarinmodified silicones for this purpose. On their own, the presence of coumarin groups converts silicone oils into thermoplastic elastomers through physical (noncovalent) cross-linking. UVirradiation permits covalent cross-linking through [2 + 2] cycloadditions and is accompanied by loss of most physical cross-links. Higher energy photons permit, in part, photoinitiated retro-cycloaddition and a subsequent decrease in covalent cross-link density. It is thus possible to tailor the physical properties of the elastomer to increase and/or decrease the modulus of the elastomer using light and to convert thermoreversible thermoplastics, by degree, into thermosets.

■

groups like azobenzene8 or ligands able to undergo photoinitated linking reactions including the thiol−ene reaction9,10 and, of course, radical polymerization,11,12 among others. A broad range of materials and applications utilizing this approach have been reported,13 and excellent reviews on this subject are available.14−17 Reversible reactions are of particular interest. For example, reversible covalent bond formation between cinnamoyl, coumarin, or other photodimerizable groups by varied wavelengths of UV light have been used to provide tunable molecular weights and polymer architectures in poly(ethylene glycol) (PEG)18 as well as facilitating applications including shape-memory materials, responsive hydrogels, and self-healing elastomers19 made from polymers such as polyacrylates,20 hyaluronic acid,21 and polyurethanes.22 In spite of the broad use of silicones in a variety of applications, there are very few examples of stimuli-responsive silicone polymers. Lewis acid/Lewis base interactions between silicone boronates and aminoalkylsilicones are reversible at elevated temperatures and in the presence of simple amines.23 Heat-responsive silicone elastomers (thermoplastics) result when silicones bear organic groups that self-assemble. For example, block copolymers of silicone/urethanes are thermoplastic.24 Previously, we reported that coumarin-modified silicone polymers possess the ability to self-assemble.25 Thermoplastic silicone elastomers result, and the physical cross-link density was found to be directly proportional to the concentration of coumarin groups on the polymer backbone. The presence of coumarin groups on the silicone means these

INTRODUCTION The wide utility of polydimethylsiloxane (PDMS) elastomers could be further enhanced if it was possible to tune their properties after curing: silicones are normally thermoset materials. While the ability to dial in the elastomer properties by stoichiometric control of the starting materials is one of the advantages of silicones, the ability to adjust properties after initial cure would be beneficial. Such an approach has been used, for example, to change the refractive index of intraocular lenses using light after implantation.1,2 Stimuli-responsive materials have become a popular research focus over the past decade due to the ability to use external sources such as heat or light to induce structural or property changes on demand within the material. As a consequence of precise, localized control over polymer morphology, the properties of lightresponsive polymers, particularly those that allow for reversible transitions, can be tuned at will. Silicone elastomers are typically made commercially using radical, hydrosilylation, or condensation cure (Figure 1A−C).3 The former processes involve cure via organic spacers, while the latter give pure siloxane cross-linkers: different thermal cure strategies using the Piers−Rubinsztajn reaction also lead to siloxane bridges.4 Silicone elastomers may also be formed through the polymerization of silicone polymers with pendant organic monomers. For example, a rapid photocuring process was described by Crivello,5,6 in which photoacids initiate epoxide ring-opening polymerization, and the photocure of acrylic-modified silicones is well-known (Figure 1D,E).7 Many types of polymers can be made into photoresponsive, covalently bound networks and materials through introduction of an appropriate functionality during polymer or elastomer synthesis. Such functionalities could include photoswitching © XXXX American Chemical Society

Received: May 20, 2015 Revised: August 20, 2015

A

DOI: 10.1021/acs.macromol.5b01085 Macromolecules XXXX, XXX, XXX−XXX

Article


Macromolecules

Figure 1. Selected cure strategies for silicones.

Figure 2. Photodimerization/cleavage of coumarin. R = silicone oligomer or polymer.

undergoes a reversible [2 + 2] photocycloaddition with λ > 300 nm to give a cyclobutane that photocleaves with λ < 300 nm

materials are also photoresponsive. In principle, any changes resulting from photodimerization are reversible: coumarin itself B


Article


Macromolecules

Figure 3. Synthesis of C-PDMS and SC-PDMS. (A) Synthesis of azide-functional silicone; (i) NaN3, Bu4NN3, THF, RT, 2−48 h, (ii) THF removed, filtered through neutral alumina with Et2O. (B) Huisgen reaction to create C-PDMS containing different concentrations of coumarin; alkyne-functionalized coumarin, 50 °C, 4 days. (C) Thermal cross-linking to form SC-PDMS-1.5 using an α,ω-dialkynylsilicone and alkynefunctionalized coumarin; 50 °C, 4 days. Note: an approximately 75:25 mixture of 1,4- and 1,5-triazole isomers is produced in the Huisgen cyclization; only the 1,4-isomers are shown.

(Figure 2).26 These materials should be amenable to photoinduced hardness patterning in 3D and/or following physical degradation/cracking should be amenable to photohealing. We have previously developed strategies to control the number of coumarin functional groups that can be introduced along the PDMS backbone (Figure 3).25 In this paper, we examine the ability to manipulate the cross-link densityboth through 1:1 coumarin:coumarin physical interactions and through covalent bonds that result from photodimerization of the coumarin groupsturning a thermoplastic into a thermoset elastomer. In addition, the ability to initiate retro-cycloaddition, therefore partly reducing the covalent cross-link density, is demonstrated.

viscosity increases of several orders of magnitude were observed, as we previously demonstrated.25 The compounds were shown to be thermoplastic elastomers because 1:1 coumarin:coumarin π−π interactions act as cross-links. These interactions can be overcome at elevated temperatures: heating−cooling cycles of the non-cross-linked materials led to circular plots of modulus against temperature (e.g., CPDMS-14, Figure 4A) because of the reorganization of selfassembled coumarin groups that occurred as temperatures were increased or decreased. Coumarin-functionalized PDMS were photodimerized on a photorheometer using a 365 nm light source (C-PDMS → PCPDMS). Optimal temperatures for the photocuring reaction were initially obtained through preliminary temperature sweep tests of each material, such that each of the polymers would cure starting from approximately the same initial viscosity, which ranged between 195 and 294 Pa·s (C-PDMS-3: 40 °C; C-PDMS-7: 62 °C; C-PDMS-11: 74 °C; C-PDMS-14: 74 °C): the final moduli of the PC-PDMS materials were more easily compared when all four compounds had the same initial viscosity. Changes in modulus during cure were followed over time of UV irradiation (365 nm, mW/cm2). The mechanical properties of the photo-cross-linked elastomers were quite different from the analogous physically cross-linked precursors (Figure 4B,E). During irradiation in the melt, each polymer became more elastomeric, as seen from changes in the storage modulus G′, which crosses over the loss modulus, G″ (Figure 5A,D).

■

RESULTS A variety of coumarin-functional PDMS polymers (C-PDMS) were prepared using the Huisgen−1,3-thermal-cycloaddition reaction between alkyne-functionalized 7-hydroxycoumarin and various azide-functional PDMS (∼9 kDa, Figure 3A) to give coumarinylpropylmethyl-co-dimethylsilicones (copolymers with 3, 7, 11, and 14% coumarin C-PDMS-3, -7, etc., Figure 3B).25 For one of the samples SC-PDMS-1.5, the azide backbone of C-PDMS-3 was partly covalently cross-linked, irreversibly, using an alkyne-terminated PDMS of approximately 1 kDa; coumarin groups were introduced onto the remaining azido groups (y = 0.11, Figure 3C). The introduction of coumarin groups on the backbone led to surprising changes in the physical properties of the silicone: C


Article


Macromolecules

Figure 4. Thermal cycling rheology data for (A) C-PDMS-14 (before irradiation) and after the C-PDMS samples had been irradiated to give (B) PC-PDMS-3, (C) PC-PDMS-7, (D) PC-PDMS-11, (E) PC-PDMS-14, and (F) SC-PDMS-1.5 (bottom) and PSC-PDMS-1.5 (top).

Figure 5. Irradiation of C-PDMS (in the melt) → PC-PDMS: (A) PC-PDMS-3 at 40 °C, (B) PC-PDMS-7 at 62 °C, (C) PC-PDMS-11 at 74 °C, and (D) PC-PDMS-14 at 74 °C.

off to cure C-PDMS-3 (in the melt) → PC-PDMS-3 (Figure 6). Identical behavior to the continuous photodimerization study (Figure 5A) was observed, except that the viscosity

In order to confirm that the modulus increase with exposure to 365 nm light was a direct consequence of photodimerization, rather than thermal side reactions, the lamp was cycled on and D


Article

Macromolecules

consequence (Figure 4F). The total number of cross-links in PSC-PDMS-1.5 should be approximately the same as in PCPDMS-3. However, the longer spacers initially used to partly cross-link the material lead to an overall lower cross-link density that is manifested in a lower modulus (Figure 4B vs 4F-top). Instron studies after photolysis were undertaken to characterize the elastomer mechanical properties after removal from the rheometer (Figure 7, PC-PDMS-7 could not be recovered from


Figure 6. C-PDMS-3 irradiation to give PC-PDMS-3 with lamp on/ off cycling. In each cycle the lamp is on for 18 min (gray), then off for 12 min, on for 18 min, etc.

plateaued immediately after each 18 min irradiation. This behavior is consistent with photodimerization as the only cause of the changes in modulus. Once a plateau in viscosity/modulus was reached on the rheometer, that is once photodimerization had reached a maximum, the PC-PDMS samples were tested using thermal cycling to confirm cure completion (Figure 4B,E). As seen by comparing C-PDMS-14 (Figure 4A) with PC-PDMS-14 (Figure 4E), after irradiation there is essentially no change in modulus as a function of temperature: once covalently crosslinked, little further change can occur. That is the thermal response of cured materials followed a straight line with temperature, showing that after photo-cross-linking there are insufficient free coumarin groups left to permit significant physical coumarin/coumarin cross-links, or the geometric constraints of the network preclude additional physical coumarin dimers. The exception to this was PC-PDMS-11, the only sample that still underwent small physical changes during thermal cycling even after 18 h of UV exposure. With this particular sample, it is possible that some coumarin groups could not find the restricted orientations necessary for photodimerization but were able to physically associate due to the ways in which coumarin groups were distributed along the chain. PC-PDMS-11 was also examined using 1H NMR to determine the approximate photo-cross-linking efficiency. Based on previously reported literature studies of dimerized coumarin-functionalized poly(ethylene glycol) (PEG), new peaks from the cyclobutane ring should appear in the spectra around 3.8−4.8 ppm as well as significant shifting of the coumarin protons to 6.1−7.0 ppm.18 Solid-state proton NMR experiments demonstrated, for PC-PDMS-11 and PC-PDMS3, that approximately 66% (Supporting Information) and 88% of the grafted coumarin moieties were photodimerized, respectively. This corresponds well with what was observed from rheology (see next section). The relative efficiency of photo-cross-linking should depend on the degrees of freedom of the coumarin groups, which will diminish as cross-link density increases. With a higher concentration of coumarin molecules on the backbone, efficient cross-linking will initially occur but will be increasingly inefficient as polymer chain mobility becomes more restricted, as is the case with PCPDMS-11. The partly covalently cross-linked polymer SC-PDMS-1.5 was also exposed to UV light to initiate formation of a second set of covalent cross-links, with the modulus increasing as a

Figure 7. Instron results of the PC-PDMS samples (based on a minimum of three repeats) and results for comparison of C-PDMS-11 and -14. The unfilled silicone control is plotted on the inset graph.

the rheometer to produce a pristine sample for testing). The photodimerized samples followed typical elastomeric tensile curves similar to commercial PDMS elastomer controls. To demonstrate this, two controls were also tested: an unfilled PDMS control (plotted on the inset graph) and a commercial sample that contained the silica fillers that are commonly used to reinforce PDMS. The former material corresponds closely to PC-PDMS-3 and shows substantially more strain than the filled analogues. As expected, these data suggested that covalent cross-links by dimerizing coumarin were as effective as traditional ethylene spacers formed by hydrosilylation in silicone rubber. When the number of cross-link sites was further increased with the presence of additional coumarin groups, e.g., PC-PDMS-11 and -14, the compounds showed a much lower strain; i.e., they were much more brittle. The strength of the photocured materials was much higher when compared to the uncured thermoplastic starting material (cf. C-PDMS-11 vs PC-PDMS-11 and C-PDMS-14 vs PCPDMS-14, Figure 7).25 C-PDMS-3 was a viscous fluid, while its photodimerized analogue PC-PDMS-3 was a very, very soft elastomer. A substantial difference can be seen between CPDMS-14 before photo-cross-linking and after. Before photodimerization, C-PDMS-14 was very brittle (with a Shore A value of ∼65), whereas the stress PC-PDMS-14 could withstand after photodimerization had more than doubled. As expected and as observed in the rheology measurements, the incorporation of covalent cross-links into the material increased the modulus of the material. Both PC-PDMS-11 and -14 exhibited higher elastic tensile moduli in comparison to their thermoplastic, nonirradiated analogues (C-PDMS-11 and -14). The photoreversibility of the system was examined in a UV− vis spectrometer using two light sources: 365 nm for dimerization and 254 nm for retro-cycloaddition. The experiment was designed such that the temperature within the spectrometer was constant at 20 °C, and gentle stirring was used to ensure even mixing of the material throughout the cuvette rather than simply where the light was directed. In the E


Article


Macromolecules

Figure 8. UV−vis of PC-PDMS-3 after 365 nm irradiation (bold solid to bold dashed): (A, C) demonstrating loss of free coumarin groups from dimerization, then (B) 254 nm irradiation (bold dashed to narrow solid) showing regeneration of coumarin groups, and (D) the composite of three processes showing changes leading to a photostationary state.

Figure 9. Covalent cross-linking involves (A) dissociation of π-stacked coumarin molecules and (B) reversible photodimerization of the coumarin molecules.

first cycle, the samplea solution of C-PDMS-3 in acetonitrilewas irradiated with 365 nm light (Figure 8: bold solid to bold dashed). Next, the retro-cycloaddition reaction was run at 254 nm light for 2 h (bold dashed to narrow solid): after the first hour further changes were negligible. The final cycle involved redimerization using 365 nm light (Figure 8: narrow solid to narrow dashed). It can be seen that recovery after each photocycle was less than 100%. This suggests the formation, eventually, of a photostationary state for the polymer. That is, the system comes to a steady state, where further physical changes due to net differences in the forward/ backward photoreactions are not observed. This is a

phenomenon that is typical when photoactive molecules are incorporated into polymeric materials.27

■

DISCUSSION Previously, we showed that alkyne coumarin groups, clicked onto linear low molecular weight PDMS chains, led to novel thermoplastic elastomers through physical coumarin dimers.25 In those studies, the thermal response on the viscoelasticity of the materials was shown to vary with the concentration of coumarin along the backbone. A higher temperature was required to achieve comparable viscosities between samples with a high concentration of coumarin groups compared to F


Article


Macromolecules

with what would be expected. PC-PDMS-14 contains double the coumarin moieties of PC-PDMS-7; however, as previously mentioned, once the covalent cross-linking begins to be established, there are insufficient degrees of freedom for a large fraction of the coumarin groups to reorient to allow for further coumarin cross-linking: only about half of the available coumarin moieties are able to photolink. PC-PDMS-11 displayed the largest degree of modulus increase, a 130% increase in modulus over PC-PDMS-7, despite having only 4% more coumarin groups along the silicone chain. This demonstrates an optimal coumarin concentration along the PDMS chain for maximum crosslinking efficiency by both physical and chemical means. There appears to be no loss in viscosity during initial irradiation, which suggests there are free coumarin moieties that are unable to physically dimerize but are in a conformation to initiate the [2 + 2] photoreaction (this is true also for PC-PDMS-14). The partly cured material SC-PDMS-1.5 did not provide, upon 365 nm light exposure, the same magnitude of modulus change as its C-PDMS-3 counterpart (Figure 4B vs 4F (before UV)). The photorheology was run at 40 °C, the same as for CPDMS-3. However, in this instance the polymer started with a lower number of coumarin physical cross-links and thus had a lower initial viscosity. The covalent cross-links initially prepared by azide−alkyne “click” chemistry with a spacer silicone chain of 1000 MW (C-PDMS-3 → SC-PDMS-1.5) did not significantly change the cross-link density because of the length of the spacer. As the physical coumarin cross-links were converted to chemical cross-links, the material quickly became more elastomeric and the storage modulus increased. The viscosity in the SC-PDMS-1.5 samples thus plateaued more quickly than the C-PDMS-3 samples. These results are interesting because they demonstrate the ability to independently utilize physical, irreversible covalent, and reversible covalent interactions to tailor the silicone elastomer properties. Thermal cycling experiments were completed on all photocross-linked samples and can be compared to previously reported thermally cycled unphotodimerized, self-associated materials.25 Below the melting temperature, which leads to loss of physical cross-linking, the cross-links are approximately equal in their ability to affect modulus. The covalent cross-links led to a slightly higher modulus than physical cross-links, as shown for C-PDMS-14 in Figure 4A vs PC-PDMS-14 Figure 5D, respectively. This further supports the proposal that 1:1 coumarin cross-links are involved in both physical and covalent cross-links: for C-PDMS-14 it was previously demonstrated by titrating with monofunctional coumarin that only two coumarin molecules are involved in a physical cross-link.25 If there were multiple coumarin groups involved in physical cross-linking, at low temperatures the initial viscosity of the physically crosslinked material would be higher than the final viscosity of the material after photodimerization. Once heated, physically cross-linked materials experienced a dramatic loss of strengththe cross-links were mostly cleaved. By contrast, thermal cycling for photodimerized PC-PDMS materials does not exhibit circular changes in the flow of the material after UV irradiation due to the presence of thermally irreversible covalent cross-links within the network (Figure 4). Instron studies further reflect the changes as C-PDMS was photodimerized to give PC-PDMS materials. The thermoplastic nature of C-PDMS-11 and -14 are shown in the curves of Figure 7. However, once photodimerized, the materials exhibit a typical thermoset tensile strength curve. The changes

those of lower concentration. In other words, higher coumarin backbone density gave higher physical cross-link densities. Thermal cycling was completed on all physically cross-linked materials, demonstrating the response of the various polymers to heat. All polymers exhibited reproducible changes in viscosity as the temperature was cycled; SC-PDMS-1.5 showed the least change in viscosity with temperature, which is due to the permanent cross-links incorporated into the material. Since SC-PDMS-1.5 is already cross-linked, the polymer chain movement with changing temperatures is limited, although slight cycling is noted from the effect of the coumarin selfassembly (Figure 4F). Here we examined additional changes brought through UV irradiation. The [2 + 2] photodimerization of all C-PDMS samples was explored using real-time photorheology measurements during 365 nm irradiation (Figure 3). The temperature of the dimerization experiment was different for each sample, depending on the concentration of coumarin, in order to maintain approximately constant initial moduli across all samples. Except for PC-PDMS-14, starting from the melt polymers with higher coumarin backbone concentrations demonstrated an increase in G′ after UV exposure: the modulus of PC-PDMS-14 did not exceed that reached by PC-PDMS-11. Increases in modulus are a consequence of covalent cross-linking through [2 + 2]-cycloaddition reactions: the deviation by PC-PDMS-14 is likely a consequence of too high a density of coumarin moieties. At these higher loadings, we rationalize that complete dimerization of coumarin along the polymer backbone is not possible due to the conformational restraints placed within the material during cross-linking: there are insufficient degrees of freedom to permit more than a fraction of coumarin to attain the geometry necessary for photocycloaddition or physical 1:1 coumarin:coumarin association. The polymers possessing lower coumarin concentrations along the backbone, PC-PDMS-3 and PC-PDMS-7, exhibited an unexpected initial decrease in viscosity when the lamp was turned on at 2 min (the initial 2 min of measurement without irradiation were to ensure no thermal processes were occurring Figure 5A,B). We postulate this decrease in viscosity is due to the conformational changes that occur when photodimerization begins, which first involves rearrangement of the coumarin groups away from the physical dimer structure (Figure 9A).28 When the polymer is initially placed on the rheometer, the coumarin molecules are stacked in a favorable head-to-tail conformation that leads to the viscosity increase previously reported. When the lamp is turned on, the associated coumarin groups must first disassembleleading to lower viscosity materialsprior to the orientation permitting them to undergo cycloaddition. Note that thermal association of the coumarin molecules was found to arise from head to tail, π-stacking arrangements in model studies.29 By contrast, the photodimer can arise from either head-to-tail or head-to-head interactions (Figure 9B).30 Photocuring of all of the compounds tested led to increases in moduli. PC-PDMS-3 had the lowest final viscosity, which is expected as it contains the lowest concentration of coumarin and therefore the lowest possible concentration of coumarinderived photo-cross-links. PC-PDMS-7 and PC-PDMS-14 increased by the same degree in modulus, just over double the final modulus for PC-PDMS-3. Compared to PC-PDMS-3, PC-PDMS-7 contains about double the percentage of coumarin moieties, and so the increase in modulus follows G


Article


Macromolecules

geometric requirements, as can be seen from the model structures in Figure 9. The modulus at lower temperatures can be manipulated using both physical and photochemical cross-links, while at elevated temperatures the physical cross-links vanish. The conversion of a selected fraction of physical into covalent crosslinks using light allows one to tune the mechanical properties of the silicone elastomer at various temperatures and to do so, to a degree, reversibly. Efforts to combine the processes with patterning will form the basis of future reports.

to PC-PDMS-11 and -14 show an increase in stress at break over their physically cross-linked analogues, which is expected considering the incorporation of stronger chemical cross-links within the material. PC-PDMS-3 exhibits a tensile strength that is very similar to the unfilled-PDMS control, which demonstrates that there is a concentration of coumarin photo-cross-links that can produce materials with a similar tensile strength as silicones containing alkyl cross-links. These tensile strength results demonstrate the utility of coumarin-functionalized silicones in the formation of covalent adaptable networks (CANs), a category of network-forming polymers that combine the desirable processability of thermoplastics (e.g., recyclability, remoldability, etc.) with the added structural resilience of thermosets through reversible or dynamic covalent bond formation.11 The studies above show it is possible to combine thermal and photoinduced cross-links to tune the properties of the materials. The final physical parameters depend upon the total coumarin concentration along the polymer chain, and the number physical and covalent cross-links, which also depend on the available orientation of coumarin molecules: some molecules will be conformationally stranded and unable to participate in either type of cross-link. Thus, starting from the same linear silicone, polymers from oils to soft, then hard, elastomers and resins can easily be prepared and then modified photolytically. Coumarin-containing polystyrenes31 and poly(ethylene glycol)s32 have also been studied and demonstrated physical behaviors similar to C-PDMS compounds. Fu et al. demonstrated small differences between linear polystyrene with pendant coumarin groups (a homopolymer prepared via RAFT polymerization) and a polyester dendrimer capped with coumarin and found that the solution-state photo-cross-linking efficiency (71% at a UVA dose of 24.8 J cm−2) and photoreversibility (31−84% within three cycles) of the linear polystyrene photopolymer were less than those of the dendrimer.31 Yamamoto et al. had similar results for poly(ethylene glycol)s bearing pendant coumarin groups (prepared through polycondensation of PEG with the diacyl chloride of a coumarin derivative), demonstrating solid-state cross-linking efficiency of less than 100% and a maximum of 75% photocleavage after extended periods of irradiation at 254 nm.32 A notable similarity between the two pendant photoresponsive polymers discussed in these two literature reports is the preference for photoreactions in solution over the solid state. It was found that the increased flexibility of polymer chains led to faster rates of photodimerization. This is one of the desirable benefits of the physical properties possessed by CPDMS photopolymers: because of the highly flexible silicone backbone, the polymers are more mobile and can readily undergo facile photochemistry as highly viscous materials over a broad range of temperatures, unlike more rigid polyethers and hydrocarbon polymers. The key advantage of these coumarin-modified silicone materials is the ability to incorporate specific concentrations of physical and covalent cross-links into the elastomer: both types of cross-link may be reversed either thermally or photochemically, although the efficiency of photoreversion is less than 100%. The cross-links arise from different modalities of 1:1 coumarin:coumarin interactions in both cases. The physical interaction is best understood as a head-to-tail π-stacking interaction.25 By contrast, both head-to-head and head-to-tail dimerization is possible for the [2 + 2] photocyclodimerization.33 These two dimerization modalities have different

■

CONCLUSIONS Coumarin-modified silicones provide two distinct methods to control elastomeric properties. The concentration of coumarin incorporated along the silicone backbone controls the initial viscosity and thermal properties of the material. In addition, the photodimerization of coumarin within PDMS allows the thermoplastic silicone polymers to be converted by light into thermosets of the desired modulus. Covalent cross-link density can be controlled by the total density of coumarin groups along the chain and by the length of photolysis time at a given intensity. To a degree, the photo-cross-linking process is reversible such that the elastomeric modulus can be partly reduced by photolysis at shorter wavelengths.

■

EXPERIMENTAL SECTION

Materials. Sylgard 184 was purchased from Dow Corning as a twopart kit. Tetrahydrofuran and toluene were purchased from Caledon and dried before use over an activated alumina column. Ethanol, acetonitrile, chloroform, and hexanes were purchased from Caledon and used as received. Deuterated chloroform (99.8%) was obtained from Cambridge Isotope Laboratories. Glycidyl ether-terminated PDMS (DMS-E12, 956 g/mol by 1H NMR analysis) was purchased from Gelest. Propiolic acid (95%) was purchased from Aldrich.

All materials for the synthesis of C-PDMS polymers were obtained, synthesized, and characterized following previously reported experimental procedures.25 Methods. Nuclear Magnetic Resonance Analysis. 1H and 13C NMR spectra were collected on a Bruker Avance 600 MHz or a 500 MHz spectrometer using chloroform-d as solvent. The proton impurity of the deuterated solvent was used as a reference for 1H NMR spectra (chloroform = 7.24 ppm). Rheological Analysis. Rheometry was performed on an ARES 3ARES-9A rheometer (TA Instruments, USA) with parallel plate geometry and a 0.3 mm gap.34 The top plate was a 20 mm quartz plate with a Peltier plate on the bottom. A dynamic temperature step test was performed at oscillatory frequencies of 10 rad s−1 and a strain of 1% and a temperature step time of 3 °C/min to obtain data for the change in dynamic viscosity as a function of temperature. Photo-crosslinking reactions were performed using oscillatory frequencies of 10 rad s−1 and a strain of 1%, with a 365 nm light source at an intensity of 96 mW/cm2 at temperatures that gave a starting modulus of ∼1000 Pa. The UV light source was an EXFO Acticure 4000. Reversible UV Studies. The UV studies were performed using a Varian Cary 50 Bio UV−vis spectrometer with a single cell Peltier accessory to ensure even mixing and controlled temperature during irradiation. The 254 nm light source was a UVP PenRay. Experiments were run in acetonitrile at a concentration of 10−4 M. Each irradiation experiment was run for 1 h, followed by an extra 15−60 min to ensure complete reaction. Instron Analysis. The tensile tests were performed on an Instron 3366 (Table model), USA, tensile testing machine at room temperature with a 50 N load cell. The specimen gauge length was H


Article


Macromolecules

(5) Jang, M.; Crivello, J. V. J. Polym. Sci., Part A: Polym. Chem. 2003, 41 (19), 3056−3073. (6) Putzien, S.; Louis, E.; Nuyken, O.; Crivello, J. V.; Kühn, F. E. J. Appl. Polym. Sci. 2012, 126 (4), 1188−1197. (7) Müller, U.; Timpe, H. J.; Neuenfeld, J. Eur. Polym. J. 1991, 27 (7), 621−625. (8) Barrett, C. J.; Mamiya, J.-i.; Yager, K. G.; Ikeda, T. Soft Matter 2007, 3 (10), 1249−1261. (9) Lowe, A. B. Polym. Chem. 2014, 5 (17), 4820−4870. (10) Fouassier, J. P.; Lalevee, J. Polymers 2014, 6 (10), 2588−2610. (11) Chatani, S.; Kloxin, C. J.; Bowman, C. N. Polym. Chem. 2014, 5 (7), 2187−2201. (12) Zhu, C.; Ninh, C.; Bettinger, C. J. Biomacromolecules 2014, 15 (10), 3474−3494. (13) Cardenas-Daw, C.; Kroeger, A.; Schaertl, W.; Froimowicz, P.; Landfester, K. Macromol. Chem. Phys. 2012, 213 (2), 144−156. (14) Zhu, M.; Rong, M. Z.; Zhang, M. Q. Polym. Int. 2014, 63 (10), 1741−1749. (15) Ma, X.; Tian, H. Acc. Chem. Res. 2014, 47 (7), 1971−1981. (16) Wischerhoff, E.; Badi, N.; Laschewsky, A.; Lutz, J.-F. Smart Polymer Surfaces: Concepts and Applications in Biosciences. In Bioactive Surfaces; Borner, H. G., Lutz, J. F., Eds.; Springer: Berlin, 2011; Vol. 240, pp 1−33. (17) Qing, G.; Li, M.; Deng, L.; Lv, Z.; Ding, P.; Sun, T. Mini-Rev. Med. Chem. 2013, 13 (9), 1369−1380. (18) Trenor, S. R.; Long, T. E.; Love, B. J. Macromol. Chem. Phys. 2004, 205 (6), 715−723. (19) Kaur, G.; Johnston, P.; Saito, K. Polym. Chem. 2014, 5 (7), 2171−2186. (20) Lendlein, A.; Jiang, H.; Junger, O.; Langer, R. Nature 2005, 434 (7035), 879−882. (21) Miyamoto, K.; Sasaki, M.; Minamisawa, Y.; Kurahashi, Y.; Kano, H.; Ishikawa, S.-i. J. Biomed. Mater. Res. 2004, 70A (4), 550−559. (22) Ling, J.; Rong, M. Z.; Zhang, M. Q. J. Mater. Chem. 2011, 21 (45), 18373−18380. (23) Dodge, L.; Chen, Y.; Brook, M. A. Chem. - Eur. J. 2014, 20 (30), 9349−9356. (24) Ward, R. S. Medical Device & Diagnostic Industry Magazine, 2000, http://www.mddionline.com/article/thermoplastic-siliconeurethane-copolymers-new-class-biomedical-elastomers, accessed Feb 8, 2015. (25) Fawcett, A. S.; Brook, M. A. Macromolecules 2014, 47 (5), 1656−1663. (26) Kehrloesser, D.; Baumann, R. P.; Kim, H. C.; Hampp, N. Langmuir 2011, 27 (7), 4149−4155. (27) Xiao, S.; Yi, T.; Li, F.; Huang, C. Tetrahedron Lett. 2005, 46 (52), 9009−9012. (28) Yang, R.; Peng, S.; Wan, W.; Hughes, T. C. J. Mater. Chem. C 2014, 2 (43), 9122−9131. (29) Seth, S. K.; Sarkar, D.; Jana, A. D.; Kar, T. Cryst. Growth Des. 2011, 11 (11), 4837−4849. (30) Chen, Y.; Jean, C. S. J. Appl. Polym. Sci. 1997, 64 (9), 1759− 1768. (31) Fu, Q.; Cheng, L.; Zhang, Y.; Shi, W. Polymer 2008, 49 (23), 4981−4988. (32) Nagata, M.; Yamamoto, Y. React. Funct. Polym. 2008, 68 (5), 915−921. (33) Fleming, I. Molecular Orbitals and Organic Chemical Reactions; Wiley: New York, 2009. (34) Hao, X.; Jeffery, J. L.; Wilkie, J. S.; Meijs, G. F.; Clayton, A. B.; Watling, J. D.; Ho, A.; Fernandez, V.; Acosta, C.; Yamamoto, H.; Aly, M. G. M.; Parel, J.-M.; Hughes, T. C. Biomaterials 2010, 31, 8153− 8163.

8.00 mm and a crosshead speed of 5 mm/min was used. PC-PDMS samples were removed from the rheometer and punched into Instron testing strips following procedures used for contact lens tensile testing. All reported mechanical properties are based on an average of a minimum of three to five specimens. Synthesis of Propiolate-Terminated PDMS. Glycidyletherterminated PDMS (5.0 g, 5.1 mmol) was combined with propiolic acid (1.42 g, 20 mmol) in THF (10 mL) and stirred at 50 °C over 4 days. The reaction was monitored by NMR, and on completion any remaining acid and solvent were removed in vacuo. 1H NMR (CDCl3, 600 MHz): δ 4.21 (t, 2 H, J = 9.0 Hz), 3.53−3.42 (m, 8 H), 2.96 (s, 2 H), 1.76−1.74 (m, 2 H), 1.69−1.65 (m, 2 H), 1.62−1.58 (m, 4 H), 0.52−0.50 (m, 4 H), 0.07−0.02 (m, 64 H). 13C NMR: (CDCl3, 600 MHz): δ 155.10, 75.98, 74.56, 70.96, 69.69, 68.49, 66.26, 23.32, 14.14, 1.28, 0.20. Synthesis of SC-PDMS-1.5. Azide-functional silicone (∼3% azidoalkylsilicone (Me2SiO)7(ClCH2CH2CH2(Me)SiO)0.78(N3CH2CH2CH2(Me)SiO)0.22), 5.0 g, 11.0 mmol) was combined with propargyl coumarin (1.19 g, 5.5 mmol) and propiolate-terminated PDMS (3.5 g, 2.7 mmol) with THF (10 mL) and stirred at 50 °C for 4 days. The reaction was monitored by NMR, and on completion the THF was removed using a rotary evaporator. 1 H NMR (CDCl3, 500 MHz): δ 8.36−8.25 (m, 0.11 H), 8.13−8.05 (m, 0.11 H), 7.36 (d, 0.11 H, J = 9.5 Hz), 7.54 (dd, 0.11 H, J = 9.0 Hz, J3 = 3.0 Hz), 7.28 (s, 0.11 H), 7.22−7.18 (m, 0.11 H), 6.42 (d, 0.11 H, J = 9.5 Hz), 4.63−4.43 (m, 0.22 H), 4.38 (t, 0.22 H, J = 7.0 Hz), 3.52− 3.47 (m, 2.22 H), 2.87 (d, 0.11 H, J = 7.0 Hz), 2.07−1.91 (m, 0.22 H), 1.88−1.78 (m, 1.78 H), 1.66−1.55 (m, 0.22 H), 0.65−0.60 (m, 1.56 H), 055−0.50 (m, 0.66 H), 0.13−0.02 (m, 32 H).

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.5b01085. 1 H NMR showing the conversion of C-PDMS-11 → PCPDMS-11 (66% conversion) (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail [email protected] (M.A.B.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank the 20/20 NSERC Ophthalmic Materials Network for financial support and the Natural Sciences and Engineering Research Council (NSERC) for provision of travel funds (SNEI program) to visit CSIRO, Melbourne, Australia. We also express our gratitude to CSIRO for helpful discussions and the use of their photorheometer. We also thank Prof. Mark Andrews, McGill University, for helpful discussions. Lastly, we thank Dr. Bob Berno for his help with solid-state NMR experiments.

■

REFERENCES

(1) Traeger, J.; Heinzer, J.; Kim, H.-C.; Hampp, N. Macromol. Biosci. 2008, 8 (2), 177−183. (2) Schraub, M.; Hampp, N. Ophthalmic Technologies XXI 2011, 7885. (3) Brook, M. A. Silicones. In Silicon in Organic, Organometallic and Polymer Chemistry; Wiley: New York, 2000; pp 256−308. (4) Fawcett, A. S.; Grande, J. B.; Brook, M. A. J. Polym. Sci., Part A: Polym. Chem. 2013, 51 (3), 644−652. I


Phototunable Cross-Linked Polysiloxanes - ACS Publications

Recommend Documents