Photoinduced Mechanical Cloaking of Diarylethene‐Crosslinked Microgels

The serial connection of multiple stimuli‐responses in polymer architectures enables the logically conjunctive gating of functional material processes on demand. Here, a photoswitchable diarylethene (DAE) acts as a crosslinker in poly(N‐vinylcaprolactam) microgels and allows the light‐induced shift of the volume phase‐transition temperature (VPTT). While swollen microgels below the VPTT are susceptible to force and undergo breakage–aggregation processes, collapsed microgels above the VPTT stay intact in mechanical fields induced by ultrasonication. Within a VPTT shift regime, photoswitching of the DAE transfers microgels from the swollen to the collapsed state and thereby gates their response to force as demonstrated by the light‐gated activation of an embedded fluorogenic mechanophore. This photoinduced mechanical cloaking system operates on the polymer topology level and is thereby principally universally applicable.


Introduction
[26] However, the majority DOI: 10.1002/adma.202305845 of these microgels are contingent on responsivity to a single stimulus.Contrarily, multi-responsive microgels can show interconnected responsivity and perform multiple functions gated by different stimuli. [5,7,16,27]Such serial connection of multiple stimuli to gate a functional response increases precision as well as selectivity and introduces logical conjunctions to meet the challenges of demanding material applications.
The gating of mechanical material responses is of particular interest since it allows the control over functions, such as the rerouting of mechanical energy, [28] programming mechanical information, [29] or mechanical cloaking, [30] which has been demonstrated previously for metamaterials.[33] There, a photoinduced reaction of a diarylethene (DAE) molecule was only possible under antecedent mechanochemical activation.Conversely, Otsuka and co-workers devised a light-gated mechanochemically responsive DAE.Depending on the state of photo-isomerization, the mechanochemical reaction could be induced. [34]While metamaterials are inherently limited by feature size and their dependency on structure alone, the design and synthesis of individual multiresponsive molecules are highly specialized and tailored to the associated material system.Consequently, a universal remote control over the mechanical and mechanochemical behavior of materials at the nanoscale is highly desirable.
Here, we report a light-gated mechanical cloaking system based on microgels (Figure 1).Due to their high molar mass, loose network structure, and a high degree of chain solvation, microgels have previously been found to be very susceptible to shear forces in the swollen state [18] while immune to them when collapsed. [15,20]Moreover, photoswitchable surfactants have been incorporated non-covalently within the microgel matrix to alter the volume phase-transition temperature (VPTT) of microgels depending on the photoswitching state. [35]In this work, we present a DAE-dimethacrylate photoswitch that acts as a co-crosslinker of poly(N-vinylcaprolactam) (PVCL)-based microgels.Photocyclization of the DAE from the ring-open to the ring-closed form transfers the PVCL microgels from the swollen to the collapsed state by altering the VPTT.Thereby, a mechanical cloaking function is activated by light under otherwise identical environmental conditions.Besides dynamic light scattering (DLS), transmission electron microscopy (TEM), and atomic force microscopy (AFM), the microgels' distinct reactivity against mechanical force is monitored using a mechanofluorochromic co-crosslinker Diels-Alder adduct of 9--extended anthracene and maleimide [36,37] to quantitatively assess force-induced covalent bond scission.Since this light-induced mechanical cloaking approach operates on the polymer topology level, we believe it to be a principally universal remote control over mechanical behavior at the nanoscale.

Microgel Synthesis and Characterization
[40] Afterward, the DAE crosslinker was introduced into microgel networks of VCL monomer, 1-vinylimidazole (VIM) comonomer, and crosslinker N,N'-methylenebisacrylamide in a precipitation polymerization (Figure 1).Compared to azobenzene-based surfactants that have been employed before by Santer and co-workers, [35] DAEs were covalently incorporated and thus non-leaching, operative at low loaded DAE fractions (0.1 mol%), and thermally bistable. [41]Photoswitching of the molecular DAE (Figure S1, Supporting Information) and the DAE-crosslinked PVCL microgels (Figure 2a; Figure S2a, Supporting Information) was investigated through UV-vis spectroscopy.Upon the UV-initiated ring-closing process, a new band at ≈525 nm appeared indicating the successful transformation of the ring-open DAE (DAEo) into the ring-closed DAE (DAEc) both in solution and within the microgel.Two isosbestic points underscored the clean photoreaction.The introduction of strongly hydrophobic DAE co-crosslinkers [42] into the microgels inevitably decreased their VPTT and under-mined their colloidal stability.10 mol% comonomer VIM provided electrostatic stabilization thus compensating this effect, as visualized on DAE-free microgels with a similar composition (Figure S3a, Supporting Information).In this copolymerization system, we assumed that the DAE would be mainly distributed within the central core region of the microgels, since the polymerization of methacrylate proceeded at a higher rate than the vinylfunctionalized monomers and similar observations were made on related microgels before. [41,43,44]The centralization of the DAE crosslinkers in the core region would reduce the hydrophobichydrophobic interactions of DAE-containing dangling chain segments and thereby improve the colloidal stability.
After the UV-induced formation of DAEc, the microgels shifted their VPTT from ≈35 to 26 °C due to the likely facilitated hydrophobic interactions derived from the conjugated and planar ring-closed form. [45]In addition, the r h of the DAEc microgels was also lower compared to DAEo microgels far below the VPTT hinting toward stacking or aggregation of the planar ring-closed DAE crosslinkers.Within the VPTT region, that is, between these two temperatures, DAEc microgels were more collapsed than DAEo microgels (Figure 2b), which was the prerequisite for the light-induced mechanical cloaking at constant environmental temperature.While we found the photoswitching process to be reversible (Figure S2a, Supporting Information), the restoration of the initial r h after ring-opening from DAEc to DAEo microgels could only be observed after ≈1 month (Figure S4, Supporting Information) likely because the DAE crosslinkers were kinetically trapped in their aggregated state.

Light-Induced Mechanical Cloaking
To investigate the mechanical behavior of the microgels under shear force, microgels in both photoswitching states were subjected to ultrasound [46] at 30 °C for 20 min using an immersion probe sonicator at f = 20 kHz.The hydrodynamic radii r h were measured over the course of sonication by DLS (Figure 2c; Figure S5, Supporting Information) and indicated significantly differing behavior.The DAEo microgels showed a decreasing trend of r h within the first minute of sonication, corresponding to the loss of their outer shell (fuzzy peripheral parts of dangling chains).Upon further sonication, the consecutive aggregation of the exposed core region was expressed in a drastically increased r h spurred by depletion interactions. [18,47]After 10 min sonication, the DLS measurements did not deliver any reliable fitting results due to the high dispersity of the sample caused by aggregation.In contrast, microgels with DAEc maintained an constant size during the 20 min of sonication.TEM images underlined these results and revealed that DAEo microgels lost their fuzzy outer shell and formed aggregates after 20 min of sonication (Figure 2d,e), while DAEc microgels were mostly monodisperse before and after sonication (Figure 2f,g).
To uncover any spatially confined topological changes leading to these results, we performed Force Volume measurements of DAEo and DAEc microgels by AFM.Thereby, we investigated possible changes in the contact stiffness of the microgels and corroborated their size changes before and after photoswitching.Dilute DAEo and DAEc microgel solutions were dropcasted on polyallylamine hydrochloride (PAH)-functionalized glass substrate, rehydrated, and maintained within the VPTT shift window at 30 °C.After measuring an overview image in PeakForce Tapping mode (Figure S6, Supporting Information), 10 single microgels were randomly selected and measured for each state in the Force Volume mode.Overall, the corrected height images of the microgels (Figure 3a,b) showed the expected deformation of the microgels at the solid-liquid interface when compared to their dimensions in bulk.However, the images also revealed that the microgels in both states often are not round in the xy plane, unlike conventional microgels, [48] and show asymmetrical deformities (compare also Figures S7,S8, Supporting Information).Besides the size polydispersity of the microgels, which might be exacerbated by inhomogeneous photoswitching, the microgels also appeared to have a polydispersity in regard to the distribution of the photoswitchable moieties, which may have given rise to the asymmetrical deformations of the microgels at the solidliquid interface.
When comparing the average sizes of the microgels in the different states (Figure 3e), the DAEo microgels were ≈15 nm higher than the DAEc microgels, underlining the results obtained from DLS measurements in bulk, although the microgels within both states can vary greatly in size.The comparison of the contact stiffness profiles of the microgels in both states (Figure 3c,d) showed that there was very little difference in the measurable (non-gray) region of the profiles.Toward the center of the microgels, the contact stiffness, which can be interpreted as a measure for the polymer density, increased as expected. [48]owever, there was little to no difference between the DAEo and DAEc microgels and the microgels also showed considerable variation in their contact stiffness profiles (Figures S9,S10, Supporting Information).These measurements underlined that the ring-closing process reduced the average size of the microgels, in accordance with the DLS results, while the contact profile regarding stiffness was maintained in both DAEo and DAEc states.The complete AFM images of measured microgels can be found in the Supporting Information (Figures S6-S10, Supporting Information).

Monitoring Covalent Bond Scission
To obtain a more quantitative assessment of the mechanical cloaking effect, Diels-Alder adducts of a 9--extended anthracene and maleimide were used as optical force probe (OFP) crosslinkers to report covalent bond scission events (Scheme S2, Supporting Information). [36,37]Upon force-induced chain breakage, the OFP would undergo a retro Diels-Alder reaction and thereby turn on its cyan fluorescence.Since covalent bond scission is inevitably accompanying microgel fragmentation, this probing  method hence allowed us to assess the extent of structural damage avoided by the mechanical cloaking technique.Spectral interference of the OFP fluorescence and DAEc, such as energy transfer or emission-reabsorption processes, were excluded on control molecules in solution (Figure S11, Supporting Information). [49]The introduced OFP affected the VPTT behavior of the microgels and resulted in a smaller VPTT shift window after photoswitching (Figure S12, Supporting Information) necessitating an operating temperature of 34 °C.fluorescence became visible on the formed aggregates of DAEo microgels (Figure 4c,d).Conversely, DAEc microgels remained non-fluorescent and dispersed (Figure S13c,d, Supporting Information).These results underscored the TEM measurements depicted in Figure 2d-g with a similar degradation behavior for DAEo microgels, while DAEc microgels remained mechanically protected.Fluorescence spectroscopy emphasized these opposing responses (Figure 4e).While DAEo microgels showed a continuously increasing breakage of the integrated OFP with sonication time, the DAEc microgels reached a plateau fluorescence intensity already at or before 10 min.
We hypothesized that even in the mostly collapsed state of DAEc microgels residual dangling chains in the shell remained unprotected from the cloaking effect and were susceptible to force-induced bond scission while the more centrally localized bonds were efficiently screened from the mechanical field in the early stages of sonication (Figure 4f).After 20 min, however, the dangling chains of the DAEc microgels were mostly cleaved and no further bond scission was observed.Contrarily, the unprotected DAEo microgels continued to degrade further within their central region.This was strongly suggested by the continuously increasing fluorescence intensity in combination with a notable secondary inner filter effect (reabsorption of the first vibronic emission band at 412 nm) indicating a high local fluorophore concentration within the core region.The CLSM image in Figure 4a and the seemingly identical contact stiffness of both intact DAEo and DAEc microgels in Figure 3 corroborated this interpretation.The ratio of the emission maxima at 435 nm of both samples after 20 min sonication was ≈3:1, corresponding to a cloaking efficiency of ≈66% when measured in the amount of mechanically broken bonds.

Conclusion
We have presented a light-gated mechanical cloaking approach based on the switchable phase transition of a colloidal microgel.
Therein, a photoswitchable DAE acted as an auxiliary crosslinker shifting the VPTT of the microgel.Thus, a temperature window was created where the ring-open and ring-closed DAE microgels showed distinctively different solution behavior and thus contrasting susceptibilities to shear force.DLS and TEM measurements proved that the non-cloaked microgels underwent chain breakage of the shell part, exposed their core, and subsequently aggregated.Conversely, the cloaked microgels were mostly irresponsive to ultrasonication.These results were underpinned by AFM measurements and the additional incorporation of a mechanofluorogenic OFP crosslinker that visualized covalent bond scission events by turning on fluorescence.CLSM and fluorescence spectroscopy confirmed the light-induced cloaking mechanism and, in terms of broken bonds, returned a cloaking efficiency of ≈66%.Importantly, this light-induced mechanical cloaking approach operated on the polymer topology level thereby, at least principally, being a universal remote control over mechanical behavior at the nanoscale.Future challenges for applications of this method lie in increasing the dynamic range of the VPTT shift window while maintaining colloidal stability, increasing the cloaking efficiency, and introducing rapid reversibility not only on the photoswitching level but also for the VPTT behavior.

Figure 1 .
Figure 1.Synthesis and operating principle of mechanical cloaking with DAE-crosslinked microgels including stylized phase diagram.Non-irradiated, swollen microgels are susceptible to shear force in solution and degrade rapidly by fragmentation.Conversely, UV-irradiation of the microgels within the VPTT shift window leads to a transition from the swollen to the collapsed state, thus enabling mechanical cloaking and increasing the resistance against shear force in solution.

Figure 2 .
Figure 2. a) Photoinduced ring-closing of DAE microgels and associated UV-vis spectra in H 2 O. Ring-opening is shown in Figure S2a (Supporting Information).b) VPTT shift of DAE microgels upon ring-closing and T for ultrasonication (grey).Normalized VPTT curves are shown in Figure S2b (Supporting Information).c) Hydrodynamic radius r h of microgels containing DAEo or DAEc vs ultrasonication time at 30 °C.Individual DLS curves are shown in Figure S5 (Supporting Information).d,e) TEM images of PVCL microgels with DAEo before (d) and after (e) 20 min of ultrasonication.f,g) Microgels with DAEc before (f) and after (g) 20 min of ultrasonication at 30 °C.

Figure 3 .
Figure 3. a,b) Corrected height images of single DAEo (a) or DAEc (b) microgels, respectively, measured at 30 °C in bi-distilled water on a PAHfunctionalized glass substrate.Scale bar: 250 nm.c,d) Corresponding contact stiffness profiles through the center of the single DAEo (c) or DAEc (d) microgels depicted in (a) and (b).e) Average height profiles of single DAEo or DAEc microgels.Mean values ± SD from the mean.N = 9 randomly selected microgels for each state.
The microgel dispersions before and after sonication were characterized by confocal laser scanning microscopy (CLSM) and fluorescence spectroscopy to visualize OFP scission.CLSM provided spatially resolved fluorescence information of OFP breakage (Figure4a-d; Figure S13a-d, Supporting Information).While bright-field microscopy images revealed well-dispersed colloidal solutions and no fluorescence of both DAEo and DAEc microgels before sonication (Figure 4a,b; Figure S13a,b, Supporting Information), after 20 min of sonication
and characterizations.S.S. and W.R. contributed to the AFM measurements.S.P. optimized the synthesis of DAEs.C.Y. and Y.J. conducted parts of the syntheses, UV-vis, DLS, and sonications under the supervision of S.H.Open access funding enabled and organized by Projekt DEAL.