Modulating Non-covalent Crosslinks with Molecular Switches

1 Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL. 61801;. 2 Department of Materials Sci...
0 downloads 0 Views 622KB Size
Subscriber access provided by UNIV OF LOUISIANA

Article

Modulating Non-covalent Crosslinks with Molecular Switches Eric S. Epstein, Luca Martinetti, Ravichandran H. Kollarigowda, Olivia Carey-De La Torre, Jeffrey S. Moore, Randy H. Ewoldt, and Paul V. Braun J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 21 Jan 2019 Downloaded from http://pubs.acs.org on January 21, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Modulating Non-covalent Crosslinks with Molecular Switches Eric S. Epstein,1,2,5 Luca Martinetti,3,5 Ravichandran H. Kollarigowda,1,2,5 Olivia Carey-De La Torre,3,5 Jeffrey S. Moore,1,2,4,5 Randy H. Ewoldt,1,3,5 Paul V. Braun* 1,2,3,4,5 1

Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL 61801; 2 Department of Materials Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801; 3 Department of Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, IL61801; 4 Department of Chemistry, University of Illinois at Urbana-Champaign, Urbana, IL 61801; 5 Materials Research Laboratory, University of Illinois at Urbana-Champaign, Urbana, IL 61801. KEYWORDS. Stimuli-responsive, metal coordination, reversible crosslinking, spiropyran.

ABSTRACT: The spiropyran ligand, in conjunction with a metal ion, is shown to operate as a non-covalent reversible molecular switch. Solutions containing a spiropyran-functionalized polymer and transition metal ions underwent reversible thermallytriggered (light-triggered) transient network formation (disruption) driven by the association (dissociation) of metal-ligand crosslinks. Heat triggers metal-ion-mediated crosslinking via thermal isomerization of spiropyran to its open, merocyanine conformation, and exposure to visible light triggers dissociation of polymer crosslinks. Crosslinking is found to depend on both the valence of the ion as well as the molar ratio of spiropyran to metal salt. We envision this to be a starting point for the design of many types of reversible, stimuli-responsive polymers, utilizing the fact that spiropyrans have been shown to respond to a wide set of stimuli including heat, light, pH, and mechanical force.

1. Introduction Metal-ligand (M-L) coordination complexes have been shown to add unique mechanical functionalities to polymers through their dynamic bonding nature. An important distinction between M-L complexes and covalent bonds is that M-L complexes are generally weaker than covalent bonds, and are more likely to participate in reversible bond formation. As such, when incorporated into a polymer, M-L complexes can act as sacrificial bonds that dissipate mechanical energy, increasing the polymer’s toughness,1-3 while their reversibility means that broken bonds can reform, enabling repair after a damage event.3-6 Though M-L bonds are intrinsically reversible, few M-L systems have been shown to reversibly associate and dissociate in response to external stimuli other than mechanical force or the addition of reagents. Here, we introduce molecular switches as a motif for toggling the formation and dissociation of M-L coordination crosslinks in polymers. Specifically, we demonstrate that M-L coordination crosslinks in a spiropyran (SP) functionalized polymer solution can be modulated using heat and light as independent stimuli, resulting in reversible transformations between crosslinked and un-crosslinked states. For many years SP has been used as a photochromic dye.7 When exposed to ultraviolet (UV) light many SPs switch from a visibly transparent, ring-closed constitution (generally called the SP state) to a ring-open, merocyanine (MC) form, which strongly absorbs visible light. Upon removal of UV light or exposure to visible light, the dynamic equilibrium shifts, and MC reverts back to SP. In certain environments, such as

hydrogen-bonding media, MC becomes thermodynamically favorable, enabling thermally-triggered isomerization of SP to MC (Figure 1A).8-9 In addition to the UV and visible light, the reversible isomerization between the SP and MC is modulated by interactions with chemical species including acids and bases,10 and metal ions.11-14 As we and others have shown, when SP molecules are appropriately linked to a polymer backbone, mechanical force also triggers MC isomerization.1517 The large number of potential stimuli for modulating the dynamic equilibrium between SP and MC makes this molecular switch particularly attractive for creating stimuliresponsive materials.18 Based on previous reports that showed a 2:1 coordination between MC and transition metal ions,18-21 we hypothesize that metal salts such as cobalt [Co(II)] and copper [Cu(II)] could enable reversible crosslinking between SP-functionalized polymer chains. We note, as an interest contrast to the reversible SP chemistry discussed here, recently, Boulatov et al. demonstrated mechanically-triggered irreversible crosslinking of spirothiopyran-linked polymer chains.22

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. Spiropyran as a multi-stimuli-responsive molecular switch. (A) Hydrogen bonding stabilizes MC, making 1′,3′Dihydro-1′,3′,3′-trimethyl-6-nitrospiro[2H-1-benzopyran-2,2′-(2H)-indole] thermochromic in hydrogen-bonding solvents.8-9 (B) Schematic for the proposed mechanism for thermally-triggered crosslinking and visible-light-triggered un-crosslinking of SPlinked polymers. Metal ions stabilize MC. As a proof of concept, a concentrated propylene carbonate (PC) solution of an SP derivative methacrylate copolymer (Figure S1) and divalent transition metal ions was reversibly crosslinked by externally triggering the formation of MCmetal complexes. As illustrated by the schematic of Figure 1B, heat triggers the ring opening of SP, yielding a polymer network crosslinked with the MC-metal coordination complexes. Crosslinking is reversed by irradiating the polymer network with visible light, driving MC to release the metal ions and revert to its ring-closed SP state. This mechanism provides a unique strategy for reversibly modulating the viscoelastic properties of bulk polymer solutions using external stimuli (in this case heat and visible light), and illuminates both opportunities and challenges towards designing SP linked polymers that form non-covalent networks in response to external stimuli such as light, heat, pH changes, or mechanical force. 2. RESULTS AND DISCUSSION 2.1 Foundational Studies. Reversible crosslinking was demonstrated with a SP-functionalized methacrylate (SPMA) copolymer synthesized by reversible addition-fragmentation chain-transfer (RAFT) polymerization. RAFT polymerization was selected because it is known to yield narrow polydispersity polymers and is compatible with many functional monomers.23 Synthesis details are included in the Supporting Information (Scheme S1). Di(ethylene glycol) methyl ether methacrylate (DEGMEMA) was chosen as a comonomer to generate a polymer that is highly soluble in a variety of polar organic solvents. The final random copolymer poly(DEGMEMA–SPMA) contains ~10.5 mol% SPMA, as

determined from 1H NMR (see Figure S1 and S2), and has a number-average molecular weight Mn ≈ 26 kDa and PDI = 1.16, as determined by gel permeation chromatography (see Figure S2b). The reversible complexation of poly(DEGMEMA-SPMA) with cobalt bis(trifluoromethylsulfonyl)imide (Co(NTf2)2) was probed using UV-Vis absorption spectroscopy. Figure S4A shows UV-Vis absorption spectra of dilute solutions of poly(DEGMEMA-SPMA) in PC containing varying ratios of SP to Co(NTf2)2. The UV light exposure of metal-salt free solutions generates a large absorption band centered at λ = 564 nm, corresponding to the UV-triggered MC conformation. In the presence of Co(NTf2)2, the UV light exposure results in a broad absorption band centered at λ = 506 nm, which is characteristic of trans-MC-metal complexes.24-25 Importantly, these MC-Co complexes not only form via UV excitation, but also via thermal isomerization, a process that is accelerated by applying heat. The extent of thermal isomerization of MC-Co is strongly dependent on the concentration of poly(DEGMEMA-SPMA). Whereas the thermal equilibration of dilute solutions of SP + Co(NTf2)2[0.5 mM) results in a very small degree of MC-Co activation (Figure S3B), concentrated solutions heavily favor the formation of MC-Co complexes (Figure 2). As shown in Figure 2A, heating a 24.5 wt% polymer solution containing a 2:1 molar ratio of SP:Co(NTf2)2 results in two distinct absorption peaks centered at ~450 nm and 506 nm, respectively. The presence of two distinct peaks, both significantly blue shifted from the un-complexed MC absorption peak (inset of Figure 2B), indicates that multiple MC-Co isomers coexist under these conditions.

ACS Paragon Plus Environment

Page 2 of 9

Page 3 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Figure 2. UV-Vis absorption spectra of 24.5 wt% poly(DEGMEMA-SPMA) solution in PC containing a 2:1 ratio of SP:Co(NTf2)2. Black arrows indicate sequence of experiments. (A) Absorption spectra recorded immediately after heating at 80 °C in the dark for 1 hour. Solid curves recorded in 60 s intervals show gradual absorption decay under low intensity light from the spectrometer light source. Dashed curves recorded in 15 s intervals show rapid absorption decay under high intensity white LEDs. Inset shows this decay as a function of time (yellow shaded region indicates exposure to high intensity LEDs: ~80-90 W m-2 irradiance). (B) Spectra taken before and after 180 s UV exposure (100 W m-2 irradiance at sample surface using λ = 365 nm peak wavelength light source). Inset shows absorption curves (recorded in 36 s intervals) during UV exposure of poly(DEGMEMASPMA) in PC without metal salt. (C) Absorption spectra after heating again to 80 °C (black curve) and subsequent exposure for 180s to UV light (purple curve). The dashed green curves are the deconvolved “After 80 °C” spectrum. Several studies speculate that when multiple MC complex peaks are present, the bluer shifted peak corresponds to a cisMC-metal conformation;24, 26-27 however, in the present case, we suggest further studies are necessary to elucidate the true nature of both isomers, especially with respect to their precise stoichiometry and stability. Both peaks gradually decay under the low intensity light of the spectrometer. Exposure of the solution to high intensity white LEDs (~8090 W m-2 irradiance) results in the rapid release of Co(II) ions and a dramatic shift of the equilibrium of SP back to its ring closed state (inset of Figure 2A). After heating a second time, a nearly identical distribution of MC-Co isomers is produced (Figure 2C), indicating that this process is reversible. As a rudimentary test of the degree to which complexes form under these conditions, the solution is immediately exposed to UV light. A comparison of the relative magnitudes of the deconvolved peaks of the thermally-activated MC-Co complex peaks of Figure 2C to the UV-generated peak, indicates that while UV light shifts the distribution of MC-Co isomers, as suggested by the disappearance of the 506 nm peak, the concentration of MC-Co complexes does not increase from those that were generated via thermal equilibration (as it does in Figure S3B). Thus, we deduce from this experiment that in the absence of visible light, equilibrium heavily favors MC-Co complexation in concentrated polymer solutions (we expect near quantitative conversion of SP to MC-Co isomers under these conditions). Vial inversion tests suggest spontaneous transient network formation from concentrated (25 wt%) solutions of poly(DEGMEMA-SPMA) in PC containing 2:1 molar ratios of SP to transition metal salts (either (Co(NTf2)2 or Cu(NTf2)2) (Figure S4). Initially viscous liquids, these polymer solutions yield physically crosslinked networks within 1 hour after heating on a hot plate at 80 °C. Thermally triggered network formation also occurs at room temperature over the course of several days in the dark or under low intensity ambient light. When exposed to intense white light, these transient networks revert to viscous liquids. Whereas spirocyclic compounds are

susceptible to degradation during UV excitation28-30 previous reports have shown that neither thermal isomerization of SP31, nor visible-light induced ring closure of MC30-31 causes cyclic degradation. Thus, the two external stimuli we have chosen for this system (heat and visible light) are advantageous over UV excitation for maintaining stability. Notably, network formation does not occur with monovalent salts, such as sodium bis(trifluoromethylsulfonyl)imide (Na(NTf2)), indicating reversible crosslinking requires the presence of multivalent ions. 2.2 Reversible Crosslinking Modulated by Heat and Light. Reversible crosslinking was monitored under smallamplitude oscillatory shear deformations (Figure 3). The temperature was controlled using a peltier bottom plate, while a parallel glass disk upper geometry enabled light irradiation from above using white light (Figure 3A). Figure 3B shows the dynamic moduli of a 24.5 wt% solution of poly(DEGMEMA–SPMA) and a 2:1 ratio of SP:Co(NTf2)2 in PC tracked in real time at a constant angular frequency (ω = 0.5 rad·s–1) and strain amplitude (γ0 = 5%). A strain amplitude of 5% was chosen for all experiments, as this was well within the linear viscoelastic regime of both the crosslinked and uncrosslinked polymer solutions (Figure S12). Prior to network formation, the freshly prepared solution was exposed to white light to remove any residual MC-Co complexes formed during preparation. In its initial state under white light irradiation, the solution exhibits a loss modulus G"(ω = 0.5 rad·s–1) ≈ 0.1 Pa and a storage modulus (G'(ω)) that falls below the detection limit of the rheometer due to instrument inertia (indicated by the gray region of the plot in Figure 3).32 Therefore, at 0.5 rad·s–1 the solution is completely dominated by viscous flow. The transient network formation was initiated by heating the solution to 80 °C for 1 hour. During heating, the solution switches from a viscous to a viscoelastic state, characterized by a >10-fold increase of both G'(ω) and G”(ω) at ω = 0.5 rad·s–1. Not surprisingly, the relative magnitude of G'(ω) and G”(ω) and their temperature-induced increase were frequency dependent (not shown in Figure 3). The manifestation of measurable elasticity (G'(ω)) at a relatively low frequency (ω

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

= 0.5 rad·s–1) is consistent with the formation of a polymer network (as shown in Figure 5 and discussed in Section 2.3). Upon cooling to 20 °C, both moduli jump by roughly another order of magnitude and slowly decay. This gradual decay suggests a temperature-dependent shift in equilibrium concentration of MC.

Figure 3. (A) Rheology cell contains peltier bottom plate for temperature control and glass top plate and white LEDs for visible light exposure. (B) Storage (G'(ω), closed squares) and loss moduli (G”(ω), open squares) tracked at constant angular frequency (ω = 0.5 rad·s–1) and strain amplitude (γ0 = 5%) during heat-triggered crosslinking and light-triggered uncrosslinking. Temperature indicated by red squares (crosslinking induced by 80 °C holds). Visible light triggered un-crosslinking denoted by yellow shaded regions. Gray region of plot represents the estimated inertial limit of the rheometer.33 Dashed lines connecting data points after lapses in data acquisition represent periods during which the gap size of the parallel plates was manually adjusted to maintain a constant fill. Exposure to white light causes the crosslinks to rapidly dissociate, as indicated by the rapid decay of G'(ω) and G”(ω) to their original values prior to crosslinking. This process is modulated through several crosslinking and dissociating cycles. Hysteresis is observed with each heating and cooling cycle, as G'(ω) increases from ~0.6 Pa to ~40 Pa and G”(ω) from ~9 Pa to ~100 Pa after the first and third crosslinking sequences, respectively. In addition to elevated dynamic moduli, the transient network state appears to become more kinetically stable after the first cycle, as the decay of G'(ω) and G”(ω) at 20 °C in the dark is much less pronounced after the second and third cycles. The precise origin of this

hysteretic behavior is unclear at this time, but the result of dissociating and re-associating crosslinks is restructuring of polymer chains into a network with different relaxation dynamics (as discussed in detail below). One potential source of this hysteretic restructuring of the polymer network may be residual crosslinks after light exposure that lead to enhanced association of polymer chains in the subsequent heating step. Additionally, despite our use of PC, a low volatility solvent, solvent loss is evident in our experiments as ~3% contractions in volume, corresponding to ≤ 1 wt% increases in polymer concentration, after each cycle. These two effects will need to be decoupled to fully understand the origin of the observed hysteresis. Three control experiments were performed to confirm that reversible network formation is indeed due to MC-metal crosslinking and not a different mechanism. When the experiment shown in Figure 3 was repeated without any metal salt, no reversible changes in viscoelasticity were observed (Figure S5). Second, Co(NTf2)2 was replaced with monovalent Na(NTf2). Figure S6 shows that the poly(DEGMEMA– SPMA) solution with monovalent salt exhibits no observable change in elasticity (i.e., storage modulus) after heating at 80 °C for 2 hours (twice as long as the experiment shown in Figure 3B). Despite the fact that M-L coordination crosslinks are not expected to be formed with Na(I) ions, a small increase in G”(ω) was observed after heating. Part of this is attributed to solvent loss, but a significant fraction of the increase in G”(ω) was recovered upon exposing the solution to visible light. Heat does result in conversion of SP to MC, and one possible explanation for this reversible shift in G”(ω) is stacking of MC through π-π or dipole-dipole interactions. Molecular stacking has been observed spectroscopically in systems containing SP immobilized on inorganic surfaces,34 dissolved in polymer matrices,35 and linked to polymer backbones.36 However, for this to occur in PC, a highly polar solvent, would be surprising as it has until now been assumed to take place exclusively in non-polar media.18 An alternative explanation is that the affinity between polymer and solvent shifts when pendant SP groups on the polymer are converted to MC-Na complexes. A change in solvent affinity would change the radius of gyration of the polymer chains and consequently the viscosity of the polymer solution.37-38 Determination if the observed small reversible viscosity change is due to a change in solvent quality will require further study. Based on the near-quantitative conversion of SP to MC-Co complexes in poly(DEGMEMA-SPMA) solutions with Co(NTf2)2, we hypothesized that the ratio of SP to divalent salt in solution plays a critical role in the formation or suppression of M-L crosslinks. Specifically, we expected that an excess of metal salt would favor the formation of 1:1 MC-Co complexes and thus inhibit the formation of metal-ion-mediated crosslinks. To test this hypothesis, the rheological properties of solutions containing 3:1, 2:1 and 1:1 SP:Co(NTf2)2 were compared (Figure 4). The 3:1 system exhibits very similar frequency-dependent behavior as the 2:1 system after heating and cooling, though the dynamic moduli after crosslinking appear to be somewhat diminished (Figure 4 and Figure S7). The behavior of the 1:1 system, on the other hand, differs drastically, and strongly supports our hypothesis, as G'(ω) remains below the inertial limit of the instrument during the entire course of the

ACS Paragon Plus Environment

Page 4 of 9

Page 5 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society experiment, and changes in G”(ω) are 1 to 2 orders of magnitude smaller than those observed in the 2:1 and 3:1 SP:Co(NTf2)2 samples. In this regard, the behavior of the system with 1:1 SP:Co(NTf2)2 is strikingly similar to the Na(NTf2) control case, as both solutions remain viscous liquids after heating.

Figure 4. Comparison of first (A) and second (B) cycles for 1:1, 2:1 and 3:1 SP:Co(II). Whereas drastic, reversible changes in viscoelasticity were observed with 2:1 and 3:1 SP:Co(II), no measurable change in elasticity was observed with 1:1 stoichiometry G'(ω) remained well below the inertial limit during the entire experiment). It is hypothesized that few if any MC-Co(II) crosslinks form in the 1:1 solution. The small, reversible change in G″(ω) is suggested to be due to molecular stacking of MC molecules rather than metal coordination. The data acquisition times of the 2:1 and 1:1 data are adjusted slightly to directly overlap the 3:1 sample data. Yellow shaded regions represent times at which solutions were exposed to visible light. Gray shaded regions represent the inertial limit of the instrument. All polymer solutions are 24.5 ± 0.3 wt% poly(DEGMEMA–SPMA) in PC. Furthermore, the decay of G″(ω) during light exposure for both samples fits very well to a bi-exponential kinetic model (Figure S8), whereas both the growth and decay of G″(ω) with 2:1 and 3:1 SPMA:Co(NTf2)2 deviate significantly from simple first-order kinetics (Figure S9), suggesting that the same general mechanism accounts for the reversible G″(ω) shift for both the Na(NTf2) and 1:1 SP:Co(NTf2)2 systems. With Na(NTf2) only 1:1 complexes form due to the monovalent nature of the salt, whereas with 1:1 Co(NTf2)2, the formation of 2:1 (MC)2Co coordination crosslinks is suppressed. We have thus differentiated two mechanisms at play for reversible changes in viscoelasticity of poly(DEGMEMA-SPMA)/metal ion solutions. Whether the reversible changes in viscoelasticity of concentrated solutions of poly(DEGMEMA-SPMA) and metal salts are dominated by moderate shifts in polymer-solvent affinity or metal-ionmediated crosslinking, depends both on the valence of the ion and the SP to metal salt molar ratio. 2.3 Estimating the degree of crosslinking. To reveal a more complete picture of the viscoelastic properties of the polymer solution containing 2:1 SP:Co(NTf2)2, frequency sweeps were performed prior to heating, after heat-triggered network formation, and after light-triggered network dissociation (Figure 5). The specific times at which these frequency sweeps were acquired are denoted by “ω” in the plot of Figure 3B. Figure 5A shows that the solution in its initial state is a viscous liquid over the entire experimental range of deformation timescales, having a loss viscosity (η′(ω) ≡ G″(ω)/ω) of ~ 0.23 Pa·s. After crosslinking and cooling back to 20°C, the material clearly exhibits viscoelastic

behavior. Upon exposure to visible light, the solution returns to a purely viscous state with a dynamic viscosity that is nearly identical to its original value. However, η′(ω) nearly doubles after the solution is subject to a second round of crosslinking/un-crosslinking. Figure 5 B–D show the frequency-dependent dynamic moduli at 20 °C after the first, second and third crosslinking sequences, respectively. In all three cases, the material exhibits liquid-like terminal behavior due to the transient nature of the crosslinks.

Figure 5. (A) Frequency sweeps show reversible changes in dynamic viscosity (η′(ω)) of poly(DEGMEMA–SPMA) with 2:1 SP:Co(NTf2)2 during the three crosslinking and uncrosslinking cycles shown in Figure 3. Dynamic moduli of reversible networks at 20 °C after the (B) first, (C) second, and (D) third crosslinking sequences. Blue and red lines are numerical fits of a continuous distribution of relaxation times derived from the log-normal distribution, with parameters (τp, Hp, σ);39-40 values indicated by black circles, the horizontal bar 1  is  p e . Gray regions of plots denote the estimated inertial

limit of the rheometer. Hysteresis is also evident. In each successive cycle the dynamic moduli shift to higher values at all frequencies, G″(ω) becomes less dominant (i.e., the material response becomes more elastic), and the crossover of G'(ω) and G″(ω) occurs at lower frequencies (i.e., the network exhibits slower terminal relaxation dynamics). To estimate the percentage of SP units that are elastically active after each crosslinking cycle, the data of Figure 5 B–D were modeled using a continuous relaxation spectrum H() derived from the log-normal distribution.41 The motivation behind our choice was twofold. First, a log-normal distribution for the terminal spectrum of transient networks can be rationalized.39-40 The average lifetime () of the interchain non-covalent associations can be thought of as the mean time for an association bond to break, which is an activated process characterized by a free energy barrier of height (where kBT is the thermal energy).42-46 Therefore,

according

to

Kramer’s

theory,47

.

Assuming a Gaussian distribution of energy barriers yields a log-normal distribution of escape times.48 Second, the lognormal distribution is the simplest distribution that can

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

reasonably mimic the behavior of transient poly(DEGMEMA– SPMA) networks with the fewest adjustable parameters: Hp, the peak of the relaxation spectrum, τp, the relaxation time at Hp, and σ, the standard deviation of the log-normal distribution of relaxation times (Table S1 and Figure S10). This continuous relaxation spectrum modeling approach allowed us to estimate both the plateau modulus, G0, and the longest characteristic timescale of the network, τw (see Supporting Information for details of the computation).49-50 Table S2 shows the G0 and τw values computed for the Co(II)-network after each of the three curing cycles. The estimated G0 increases dramatically between the first and second cycles, and remains essentially unaltered after the third curing cycle. The terminal relaxation dynamics become increasingly slower, as indicated by a twenty-fold increase of τw after the second cycle, and a thirty-fold increase after the third cycle. Longer relaxation times and larger plateau moduli suggest greater connectivity of the polymer network (e.g., fewer dangling chains) after M-L crosslinks undergo light-triggered dissociation followed by re-association. The computed values of G0 provided a means to estimate the number density of SP monomer units that actively contribute to the elasticity of the polymer network. The degree of polymerization (DP) of the SPMA copolymer (DP ≈ 124, calculated via 1H NMR) is significantly smaller than the entanglement DP of methacrylate-based polymers containing either aliphatic side groups of similar length as DEGMEMA (e.g., poly(n-hexyl methacrylate), DP ≈ 200) or aromatic side groups (e.g., poly(4-tert-butylphenyl methacrylate), DP ≈ 320).51 Therefore, we assume that elasticity arises primarily due to M-L crosslinks rather than physical entanglements of poly(DEGMEMA–SPMA) polymer chains. Using the phantom model of rubber elasticity52-53 and assuming a network functionality f = 4 (as expected for a 2:1 MC:metal complex), the plateau modulus G0 and the number density of elastically active crosslinks (µ) are related by:54-57





G0     kBT   kBT

(1) where ν (= μ·f/2) is the number density of elastically active network strands. As shown in Table S2, ~14.8% of SPMA in the polymer solution actively contributes to the elasticity of the polymer network after the third crosslinking sequence. We have preliminary evidence that the degree of crosslinking is increased by enhancing the MC-cation interaction strength by changing the metal ion. Previous work has shown that MC coordinates to Cu(II) ions much more strongly than Co(II) ions.58 As we show in the Supporting Information (Figure S11), Cu(NTf2)2-crosslinked poly(DEGMEMA–SPMA) exhibits significantly higher dynamic moduli at elevated temperature than Co(NTf2)2-networks. Unfortunately, Cu(NTf2)2-networks phase separate when cooled to room temperature (Figure S11B), preventing rigorous analysis. Similar to the Co(NTf2)2-crosslinked system, Cu(NTf2)2network formation is reversed by visible light and a homogeneous viscous solution was formed by exposing the phase separated system to visible light while mixing with the rheometer plates. Finally, to help guide future work, we suggest that a complementary strategy for improving the density of switchable crosslinks may be to tune the structure of the SP molecule itself. For example, adding certain chelating groups adjacent to the phenolate anion of MC has been shown to lead to highly stable 2:1 MC-metal complexes in dark

conditions without sacrificing reversibility under visible light.59-60 3. CONCLUSIONS We have demonstrated that polymer-tethered molecular switches drive the reversible formation of crosslinked polymer networks in response to heat and light. As shown by UV-Vis spectroscopy, when poly(DEGMEMA–SPMA) is mixed with Co(II) ions, MC-metal complexes form spontaneously at room temperature in the dark, a process which is accelerated (and likely encouraged) by heat. Exposure to visible light results in the rapid release of metal ions and ring closure of SP. The viscoelastic properties of poly(DEGMEMA-SPMA) solutions were probed under small-amplitude oscillatory shear during heat-activated crosslinking and light-activated un-crosslinking. This enabled us to capture both relatively large changes in viscoelasticity due to M-L crosslinks, as well as relatively small changes due to weaker interactions, which we propose are due to shifts in affinity between polymer and solvent. The use of molecular switches to toggle M-L bond formation and dissociation may enable greater control over both the processing and self-healing capabilities of M-L coordination polymers. For example, light may be utilized to locally transform damaged regions of a solid polymer or gel to a more fluid state in order to facilitate the dynamics of self-healing. These results also have implications towards mimicking biochemical processes in synthetic soft matter. For example, triggering MC-metal coordination with mechanical force using SP mechanophores15, 17 61-62could mimic force-triggered associations between natural proteins such as vinculin and talin.61-62 Mechanically-triggered, reversible crosslinking is a compelling opportunity, but several challenges remain, most notable of which is limiting spontaneous conversion of SP to MC.

ASSOCIATED CONTENT Supporting Information is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author * [email protected].

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Funding Sources This research was supported by the US Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences and Engineering under Award DE-FG02-07ER46471, through the Materials Research Laboratory at the University of Illinois. EE also acknowledges support of a National Science Foundation Graduate Research Fellowship under Grant No. 2012141509, as well as support of a Beckman Institute Graduate Research Fellowship through the Arnold and Mabel Beckman Foundation.

ACS Paragon Plus Environment

Page 6 of 9

Page 7 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

REFERENCES 1. 2.

3. 4.

5. 6. 7. 8.

9. 10.

11. 12. 13.

14.

15. 16.

17. 18. 19. 20.

Sun, J.-Y.; Zhao, X.; Illeperuma, W. R. K.; Chaudhuri, O.; Oh, K. H.; Mooney, D. J.; Vlassak, J. J.; Suo, Z., Highly stretchable and tough hydrogels. Nature 2012, 489, 133. Li, C.-H.; Wang, C.; Keplinger, C.; Zuo, J.-L.; Jin, L.; Sun, Y.; Zheng, P.; Cao, Y.; Lissel, F.; Linder, C.; You, X.-Z.; Bao, Z., A highly stretchable autonomous self-healing elastomer. Nature Chem. 2016, 8, 618. Li, J.; Illeperuma, W. R. K.; Suo, Z.; Vlassak, J. J., Hybrid Hydrogels with Extremely High Stiffness and Toughness. ACS Macro Lett. 2014, 3 (6), 520-523. Holten-Andersen, N.; Harrington, M. J.; Birkedal, H.; Lee, B. P.; Messersmith, P. B.; Lee, K. Y. C.; Waite, J. H., pH-induced metal-ligand cross-links inspired by mussel yield self-healing polymer networks with near-covalent elastic moduli. Proc. Natl. Acad. Sci. U. S. A. 2011, 108 (7), 2651-2655. Chen, Y.; Kushner, A. M.; Williams, G. A.; Guan, Z., Multiphase design of autonomic self-healing thermoplastic elastomers. Nature Chem. 2012, 4, 467. Mozhdehi, D.; Ayala, S.; Cromwell, O. R.; Guan, Z., SelfHealing Multiphase Polymers via Dynamic Metal–Ligand Interactions. J. Am. Chem. Soc. 2014, 136 (46), 16128-16131. Guglielmetti, R., Chapter 8 - 4n+2 Systems: Spiropyrans. In Photochromism, Dürr, H.; Bouas-Laurent, H., Eds. Elsevier Science: Amsterdam, 2003; pp 314-466. Shiraishi, Y.; Itoh, M.; Hirai, T., Thermal isomerization of spiropyran to merocyanine in aqueous media and its application to colorimetric temperature indication. Phys. Chem. Chem. Phys 2010, 12 (41), 13737-13745. Shiraishi, Y.; Inoue, T.; Sumiya, S.; Hirai, T., Entropy-Driven Thermal Isomerization of Spiropyran in Viscous Media. J. Phys. Chem. A 2011, 115 (33), 9083-9090. Wojtyk, J. T. C.; Wasey, A.; Xiao, N.-N.; Kazmaier, P. M.; Hoz, S.; Yu, C.; Lemieux, R. P.; Buncel, E., Elucidating the Mechanisms of Acidochromic Spiropyran-Merocyanine Interconversion. J. Phys. Chem. A 2007, 111 (13), 2511-2516. Taylor, L. D.; Nicholson, J.; Davis, R. B., Photochromic chelating agents. Tetrahedron Lett. 1967, 8 (17), 1585-1588. Phillips, J. P.; Mueller, A.; Przystal, F., Photochromic Chelating Agents. J. Am. Chem. Soc. 1965, 87 (17), 4020-4020. Görner, H.; Chibisov, A. K., Complexes of spiropyran-derived merocyanines with metal ions Thermally activated and lightinduced processes. J. Chem. Soc., Faraday Trans. 1998, 94 (17), 2557-2564. Chibisov, A. K.; Görner, H., Complexes of spiropyran-derived merocyanines with metal ions: relaxation kinetics, photochemistry and solvent effects. Chem. Phys. 1998, 237 (3), 425-442. Potisek, S. L.; Davis, D. A.; Sottos, N. R.; White, S. R.; Moore, J. S., Mechanophore-Linked Addition Polymers. J. Am. Chem. Soc. 2007, 129 (45), 13808-13809. Davis, D. A.; Hamilton, A.; Yang, J.; Cremar, L. D.; Van Gough, D.; Potisek, S. L.; Ong, M. T.; Braun, P. V.; Martínez, T. J.; White, S. R.; Moore, J. S.; Sottos, N. R., Force-induced activation of covalent bonds in mechanoresponsive polymeric materials. Nature 2009, 459, 68. Lee, C. K.; Davis, D. A.; White, S. R.; Moore, J. S.; Sottos, N. R.; Braun, P. V., Force-Induced Redistribution of a Chemical Equilibrium. J. Am. Chem. Soc. 2010, 132 (45), 16107-16111. Klajn, R., Spiropyran-based dynamic materials. Chem. Soc. Rev. 2014, 43 (1), 148-184. Kundu, P. K.; Olsen, G. L.; Kiss, V.; Klajn, R., Nanoporous frameworks exhibiting multiple stimuli responsiveness. Nature Comm. 2014, 5, 3588. Byrne, R. J.; Stitzel, S. E.; Diamond, D., Photo-regenerable surface with potential for optical sensing. J. Mater. Chem. 2006, 16 (14), 1332-1337.

21. Aleksandar, R.; Silvia, S.; Robert, B.; Conor, S.; King Tong, L.; Dermot, D., Photonic modulation of surface properties: a novel concept in chemical sensing. J. Phys. D: Appl. Phys. 2007, 40 (23), 7238. 22. Zhang, H.; Gao, F.; Cao, X.; Li, Y.; Xu, Y.; Weng, W.; Boulatov, R., Mechanochromism and Mechanical-ForceTriggered Cross-Linking from a Single Reactive Moiety Incorporated into Polymer Chains. Angew. Chem. Int. Ed. Engl. 2016, 55 (9), 3040-4. 23. Perrier, S., 50th Anniversary Perspective: RAFT Polymerization—A User Guide. Macromolecules 2017, 50 (19), 7433-7447. 24. Wojtyk, J. T. C.; Kazmaier, P. M.; Buncel, E., Modulation of the Spiropyran−Merocyanine Reversion via Metal-Ion Selective Complexation: Trapping of the “Transient” cis-Merocyanine. Chem. Mater. 2001, 13 (8), 2547-2551. 25. Natali, M.; Aakeröy, C.; Desper, J.; Giordani, S., The role of metal ions and counterions in the switching behavior of a carboxylic acid functionalized spiropyran. Dalton Trans. 2010, 39 (35), 8269-8277. 26. Fries, K. H.; Driskell, J. D.; Samanta, S.; Locklin, J., Spectroscopic Analysis of Metal Ion Binding in Spiropyran Containing Copolymer Thin Films. Anal. Chem. 2010, 82 (8), 3306-3314. 27. Tian, Z.; Stairs, R. A.; Wyer, M.; Mosey, N.; Dust, J. M.; Kraft, T. M.; Buncel, E., Spirooxazine to Merooxazine Interconversion in the Presence and Absence of Zinc: Approach to a Bistable Photochemical Switch. J. Phys. Chem. A 2010, 114 (44), 1190011909. 28. Sakuragi, M.; Aoki, K.; Tamaki, T.; Ichimura, K., The Role of Triplet State of Nitrospiropyran in Their Photochromic Reaction. Bull. Chem. Soc. Jpn. 1990, 63 (1), 74-79. 29. Baillet, G.; Giusti, G.; Guglielmetti, R., Comparative photodegradation study between spiro[indoline—oxazine] and spiro[indoline—pyran] derivatives in solution. J. Photochem. Photobiol. A: Chem. 1993, 70 (2), 157-161. 30. Tork, A.; Boudreault, F.; Roberge, M.; Ritcey, A. M.; Lessard, R. A.; Galstian, T. V., Photochromic behavior of spiropyran in polymer matrices. Appl. Opt. 2001, 40 (8), 1180-1186. 31. Burke, K.; Riccardi, C.; Buthelezi, T., Thermosolvatochromism of Nitrospiropyran and Merocyanine Free and Bound to Cyclodextrin. J. Phys. Chem. B 2012, 116 (8), 2483-2491. 32. Ewoldt, R. H.; Johnston, M. T.; Caretta, L. M., Experimental Challenges of Shear Rheology: How to Avoid Bad Data. In Complex Fluids in Biological Systems: Experiment, Theory, and Computation, Spagnolie, S. E., Ed. Springer New York: New York, NY, 2015; pp 207-241. 33. Chambon, F.; Winter, H. H., Linear Viscoelasticity at the Gel Point of a Crosslinking PDMS with Imbalanced Stoichiometry. J. Rheol. 1987, 31 (8), 683-697. 34. Ueda, M.; Kudo, K.; Ichimura, K., Photochromic behaviour of a spirobenzopyran chemisorbed on a colloidal silica surface. J. Mater. Chem. 1995, 5 (7), 1007-1011. 35. Eckhardt, H.; Bose, A.; Krongauz, V. A., Formation of molecular H- and J-stacks by the spiropyran-merocyanine transformation in a polymer matrix. Polymer 1987, 28 (11), 1959-1964. 36. Goldburt, E.; Shvartsman, F.; Fishman, S.; Krongauz, V., Intramolecular interactions in photochromic spiropyranmerocyanine polymers. Macromolecules 1984, 17 (6), 12251230. 37. Fox, T. G.; Flory, P. J., Intrinsic Viscosity-Molecular Weight Relationships for Polyisobutylene. J. Phys. Colloid Chem 1949, 53 (2), 197-212. 38. Flory, P. J., Statistical mechanics of chain molecules. Interscience Publishers: 1969. 39. Martinetti, L.; Soulages, J. M.; Ewoldt, R. H., Continuous relaxation spectra for constitutive models in medium-amplitude oscillatory shear. J. Rheol. 2018, 62 (5), 1271-1298. 40. Martinetti, L.; Carey-De La Torre, O.; Schweizer, K. S.; Ewoldt,

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

41. 42. 43. 44. 45. 46.

47. 48. 49. 50. 51.

52. 53. 54.

55. 56. 57. 58.

59.

60.

61.

62.

R. H., Inferring the Nonlinear Mechanisms of a Reversible Network. Macromolecules 2018, 51 (21), 8772-8789. Wiechert, E., Gesetze der elastischen Nachwirkung für constante Temperatur. Annalen der Physik 1893, 286 (10), 335-348. Tanaka, F.; Edwards, S. F., Viscoelastic properties of physically crosslinked networks. 1. Transient network theory. Macromolecules 1992, 25 (5), 1516-1523. Marrucci, G.; Bhargava, S.; Cooper, S. L., Models of shearthickening behavior in physically crosslinked networks. Macromolecules 1993, 26 (24), 6483-6488. Vaccaro, A.; Marrucci, G., A model for the nonlinear rheology of associating polymers. J. Non-Newtonian Fluid Mech. 2000, 92 (2), 261-273. Rubinstein, M.; Semenov, A. N., Dynamics of Entangled Solutions of Associating Polymers. Macromolecules 2001, 34 (4), 1058-1068. Tripathi, A.; Tam, K. C.; McKinley, G. H., Rheology and Dynamics of Associative Polymers in Shear and Extension:  Theory and Experiments. Macromolecules 2006, 39 (5), 19811999. Chandrasekhar, S., Stochastic Problems in Physics and Astronomy. Rev. Mod. Phys. 1943, 15 (1), 1-89. Plotkin, S. S.; Wolynes, P. G., Non-Markovian Configurational Diffusion and Reaction Coordinates for Protein Folding. Phys. Rev. Lett. 1998, 80 (22), 5015-5018. Tschoegl, N. W., The Phenomenological Theory of Linear Viscoelastic Behavior: An Introduction. Springer-Verlag: 1989. Graessley, W. W., Polymeric Liquids and Networks: Dynamics and Rheology. Garland Science: 2008. Tsukahara, Y.; Namba, S.-i.; Iwasa, J.; Nakano, Y.; Kaeriyama, K.; Takahashi, M., Bulk Properties of Poly(macromonomer)s of Increased Backbone and Branch Lengths. Macromolecules 2001, 34 (8), 2624-2629. James, H. M.; Guth, E., Statistical Treatment of Imperfectly Flexible Chains. J. Chem. Phys. 1943, 11 (11), 531-531. Flory, P. J.; Jr., J. R., Statistical Mechanics of Cross‐Linked Polymer Networks I. Rubberlike Elasticity. J. Chem. Phys. 1943, 11 (11), 512-520. Chompff, A. J.; Duiser, J. A., Viscoelasticity of Networks Consisting of Crosslinked or Entangled Macromolecules. I. Normal Modes and Mechanical Spectra. J. Chem. Phys. 1966, 45 (5), 1505-1514. Graessley, W. W., Statistical Mechanics of Random Coil Networks. Macromolecules 1975, 8 (2), 186-190. Graessley, W. W., Polymeric Liquids and Networks: Structure and Properties. Garland Science: 2004. Rubinstein, M.; Colby, R. H., Polymer Physics. OUP Oxford: 2003. Fries, K. H.; Driskell, J. D.; Sheppard, G. R.; Locklin, J., Fabrication of Spiropyran-Containing Thin Film Sensors Used for the Simultaneous Identification of Multiple Metal Ions. Langmuir 2011, 27 (19), 12253-12260. Zakharova, M. I.; Coudret, C.; Pimienta, V.; Micheau, J. C.; Delbaere, S.; Vermeersch, G.; Metelitsa, A. V.; Voloshin, N.; Minkin, V. I., Quantitative investigations of cation complexation of photochromic 8-benzothiazole-substituted benzopyran: towards metal-ion sensors. Photochem Photobiol Sci 2010, 9 (2), 199-207. Chernyshev, A. V.; Voloshin, N. A.; Raskita, I. M.; Metelitsa, A. V.; Minkin, V. I., Photo- and ionochromism of 5’-(4,5-diphenyl1,3-oxazol-2-yl) substituted spiro[indoline-naphthopyrans]. J. Photochem. Photobiol., A 2006, 184 (3), 289-297. del Rio, A.; Perez-Jimenez, R.; Liu, R.; Roca-Cusachs, P.; Fernandez, J. M.; Sheetz, M. P., Stretching Single Talin Rod Molecules Activates Vinculin Binding. Science 2009, 323 (5914), 638-641. Yao, M.; Goult, B. T.; Chen, H.; Cong, P.; Sheetz, M. P.; Yan, J., Mechanical activation of vinculin binding to talin locks talin in an unfolded conformation. Sci. Rep. 2014, 4, 4610.

ACS Paragon Plus Environment

Page 8 of 9

Page 9 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Table of contents graphic

ACS Paragon Plus Environment