Polymer Mechanochemistry in Microbubbles

Polymer mechanochemistry is a promising technology to convert mechanical energy into chemical functionality by breaking covalent and supramolecular bonds site‐selectively. Yet, the mechanochemical reaction rates of covalent bonds in typically used ultrasonication setups lead to reasonable conversions only after comparably long sonication times. This can be accelerated by either increasing the reactivity of the mechanoresponsive moiety or by modifying the encompassing polymer topology. Here, a microbubble system with a tailored polymer shell consisting of an N2 gas core and a mechanoresponsive disulfide‐containing polymer network is presented. It is found that the mechanochemical activation of the disulfides is greatly accelerated using these microbubbles compared to commensurate solid core particles or capsules filled with liquid. Aided by computational simulations, it is found that low shell thickness, low shell stiffness and crosslink density, and a size‐dependent eigenfrequency close to the used ultrasound frequency maximize the mechanochemical yield over the course of the sonication process.

DOI: 10.1002/adma.202305130[9] The underlying principle is the US-induced growth and collapse of microbubbles, which generates shear forces that elongate and eventually break polymer chains above a certain threshold length. [10,11]Yet, the ultrasonication of conventional linear polymer chains with a central mechanophore is an energy-intensive and slow process. [12]In typical experiments using an immersion probe sonicator at 20 kHz frequency, the mechanochemical reaction rates of covalent mechanophores lead to reasonable conversions only after tens of minutes to hours of ultrasonication.Approaches to overcome this limitation are being developed on the one hand by intrin sic modifications of the mechanophore itself through increasing its mechanochemical reactivity, e.g., with supramolecular mechanophores. [13]23][24][25] For applications using higher frequency diagnostic or therapeutic US, [26] mostly unrelated to polymer mechanochemistry, micron-sized gas bubbles are an established modality to enhance the US response for imaging, drug delivery, and other functional applications. [27,28]Besides stable cavitation, such microbubbles can burst upon US irradiation thereby decreasing the energy threshold for inertial cavitation.This effect has been observed on lipid- [29,30] as well as polymer-coated microbubbles. [31,32]We hypothesize that the bursting of such microbubbles may be exploited to accelerate the activation of mechanophores.However, microbubble fabrication faces several challenges, such as short lifetime, gas escape, and the inevitable structural collapse for diameters >1 μm, that complicate their combination with designer polymer materials. [33,34]ere we present a microbubble system with a tailored mechanoresponsive polymer shell (PMB).The PMBs consist of an N 2 gas core and a disulfide [35][36][37][38][39][40] mechanophore-crosslinked polymer shell, which is produced in a double emulsion-type microfluidic device.Upon US treatment, the PMBs burst, collide, and fracture thereby cleaving the disulfide mechanophores to generate thiols.These thiols are then detected and quantified by their reaction with Diels-Alder (DA) adducts of furylated fluorophores through Michael addition and the subsequent retro DA reaction, which releases and activates the fluorophore.The PMBs are compared to microfluidically produced solid particles and liquid-filled capsules of the same mechanophore-containing polymer confirming the contribution of the gas core to the accelerated covalent bond scission process.In addition, we investigate the relationship between mechanochemical reactivity and the mechanical properties of the PMB shells that were determined by atomic force microscopy (AFM).The mechanochemical experiments employing 20 kHz US indicate an inversely proportional relationship between stiffness and activation of disulfide mechanophores.Using computational simulations, we find that a thinner shell and a size-dependent eigenfrequency close to the used ultrasound frequency maximize the mechanochemical yield.

Microbubble Production and Characterization
The PMBs were synthesized in a microfluidic device by photopolymerization of a gas-in-oil-in-water (G/O/W) double emulsion (Figure 1a).Therefore, we manufactured a flow-focusing polydimethylsiloxane (PDMS) microfluidic master using 3D printing.This not only simplified the preparation of the microfluidic master mold to one step, but also allowed to vary the channel height, which is difficult using regular lithography techniques.The crucial part of this microfluidic device was the 7 μm wide nozzle where gas channel (N 2 ), oil phase channels (93% (v/v) poly(propylene glycol) diacrylate (PPGDA), 5% (v/v) bis(2methacryloyl) oxyethyl disulfide (DSDMA), 1% (v/v) photoinitiator ethyl phenyl(2,4,6-trimethylbenzoyl)phosphinate (TPO-L)), and aqueous phase channels (50% (v/v) glycerol, 5% (w/v) polyvinyl alcohol (PVA)) were focused to form PMBs (Figure 1b and Movie S1).A UV lamp for the initiation of the photopolymerization of the PPGDA/DSDMA shell was set up before production collection to obtain stable PMBs.Once the flow rates of each phase were established, the emulsion production in the microfluidic device proceeded automatized and unsupervised for at least 12 h promising future scale-up possibilities.
After the photopolymerization, the PMBs were collected in a glass vial and floated on the top of the solution (Figure 1c).The freshly formed PMBs were characterized by optical microscopy and no coalescence or fragmentation was observed for multiple days (Figure 1d).Thereby, the size distribution was also obtained by measuring the diameter of the PMBs to an average of 25 μm (Figure 1e).No size difference was observed compared with PMBs before UV polymerization in the microfluidic device (Figure S2, Supporting Information), thereby ruling out the preactivation of disulfide mechanophores due to decompression effects.The microfluidic production of the PMBs allowed the precise regulation of their diameter in a window from 20 to 80 μm by changing the flow rate of the outer aqueous phase (Figures S3 and S4, Supporting Information).The microscopic structure of the PMBs was investigated in detail by scanning electron mi-croscopy (SEM) revealing the morphology of air-dried PMBs (Figure 1f).The shell thickness of these 5% DSDMA-crosslinked PMBs was inferred to ≈670 nm by measuring a fractured shell region (Figure 1f, inset).
The stability of the PMBs in H 2 O was then investigated by staining the PMB shells with the fluorescent dye Nile red during production.Confocal laser scanning microscopy (CLSM) and epifluorescence microscopy were employed to visualize the shell of the PMBs and showed no obvious collapse of the PMBs underpinning their mechanical integrity (Figure 1g-i and Figure S5, Supporting Information).At the same time, this demonstrated the ability of the PMBs to load cargo molecules.Diameter measurements after 30 d further proved the excellent mechanical stability of the PMBs, which remained stable, while nonpolymerized PMBs collapsed already after 2 h (Figures S6 and S7, Supporting Information).

Microbubble Mechanochemistry
Subsequently, we explored the activation of the disulfide mechanophores upon exposure to US (Figure 2a).Our PMB system consisted exclusively of the two bifunctional crosslinkers, i.e., PPGDA and DSDMA, and hence formed random but flexible polymer network shells.The sonication of the PMBs was carried out in an ice bath to maintain a stable temperature at ≈4 °C to exclude potential heat-induced effects on the mechanochemical disulfide bond scission (Figure S8, Supporting Information).After 15 min of US application the initially well-defined PMBs (Figure 2b) showed clear evidence of shell fracture caused by cavitation-induced shear (Figure 2c,d).To visualize the hypothesized accompanying mechanochemical cleavage of disulfide bonds, we employed a thiol-selective probe (TSP) based on a DA adduct of furan-dansyl and dimethyl acetylenedicarboxylate, [21,41] the synthesis and characterization of which are described in Scheme S1 and S2, Figures S9-S13, Supporting Information.The TSP reacted with the mechanochemically generated thiol groups in a Michael addition subsequently spurring the retro DA reaction and thereby the fluorogenic release of furan-dansyl (Figure 2e).The TSP was added to the PMB suspension after 15 min US application and left for 3 d to equilibrate, whereupon a clear increase in fluorescence intensity at 530 nm was recorded (Figure 2f).To validate this result, we additionally performed liquid chromatography-mass spectrometry (LC-MS) before and after US application undoubtedly marking the emergence of furandansyl with a molar mass of 331 g mol −1 as compared to the original TSP with 473 g mol −1 with progressing ultrasonication time (Figures S14-S16, Supporting Information).
To underscore the increased mechanochemical reaction rates of PMBs compared to conventional linear polymer chains, we prepared disulfide mechanophore-centered linear poly(oligo(ethylene glycol) methyl ether methacrylate) (POEG-MEMA) chains with molar masses of 10 and 20 kDa for sonication.Neither covalent chain scission nor the activation of disulfide mechanophores were detected after 15 min of ultrasonication according to GPC and TSP measurements (Figure S17, Supporting Information).Moreover, we calculated the molar ratio of disulfide mechanophore to the polymer within linear polymers and the PMBs.The molar ratio of disulfide mechanophore to 10 and 20 kDa polymers were estimated to be 1:156 and 1:313, respectively.However, the molar ratio in PMBs was 1:28, which indicated a significantly higher mechanophore loading.
Having verified the force-induced bond scission qualitatively, we then proceeded to quantitatively assess the fraction of activated disulfides.Therefore, we first recorded a fluorescence cal-ibration curve with known thiol concentrations (0 to 100 μm and 100 to 500 μm) and a defined TSP concentration at the emission maximum of the released fluorophore at 530 nm.Although nonlinearity was expected [42] and observed for higher thiol concentration regimes, a correlation between fluorescence intensity and thiol concentration was successfully established (Figure 3a, Figures S18 and S19, Supporting Information).To obtain the maximum number of disulfide bonds within the PMBs as an upper boundary, the disulfides were chemically reduced using tris(2-carboxyethyl)phosphine hydrochloride (TCEP, Scheme S4, Supporting Information) and then quantified using the TSP method (Table S1, Supporting Information).Hereafter, 25 μm diameter PMBs were sonicated and the fluorescence intensity was recorded (Figure 3b) leading to a maximum of 12% disulfide scission relative to the overall available disulfide motifs with productive bond scission stopping after 7 min (Figure 3c).

Influence of Bubble Size
Since the size of the microbubbles affects their inertial cavitation performance, [43] we hypothesized that altering the PMB diameters might influence their mechanochemical reaction yield.Therefore, we additionally produced otherwise chemically identical 35 and 50 μm diameter PMBs (Figure 3d,g, Figures S20 and S21, Supporting Information; Movie S2-S4) and subjected these to US, analogously to the above conditions.While activa-tion was shown beyond doubt (Figure 3e,h), we found that relative disulfide scission was decreased to less than 2.5% in both cases (Figure 3f,i).Conversely, no significant size-dependence between the different PMB diameters was found upon using high frequency focused US at 0.68 MHz for 5 min with relative disulfide scission ranging between 7.8 and 9.7% (Figure S26, Supporting Information).

Influence of Internal Structure
In addition to PMB size, we investigated the influence of the polymer network particle structure on the mechanochemical performance by comparing PPGDA/DSDMA particles with solid polymer network core and PPGDA/DSDMA capsules filled with water to the PMBs.PPGDA/DSDMA solid core particles were produced in an oil-in-water (O/W) single emulsion microfluidic device and subsequently were photopolymerized by a UV LED (Figure 4a).The production process was monitored by microscopy (Figure S22a, Supporting Information; and Movie S5) and the particles were collected (Figure S22b, Supporting Information) showing average diameters of ≈25 μm calculated from a set of 50 particles (Figure 22c).
The solid core-like structure was verified by fluorescence microscopy using Nile red labeling (Figure 4b).Upon subjecting these particles to US and visualizing disulfide activation by the TSP method, we found that less than 1% of the disulfide bonds were cleaved after 15 min in the absence of a gas bubble core (Figure 4d,d).
PPGDA/DSDMA water core capsules were produced using a double-cross-shaped microfluidic device for water-in-oil-inwater (W/O/W) (Figure 5a and Figure S23a, Supporting Information).The oil phase forming the polymer shell of the capsules was identical to the PMB shells and the PPGDA/DSDMA particles (Movie S6).After photopolymerization and visual characterization (Figure S23b, Supporting Information), the diameter of the PPGDA/DSDMA capsules was determined to ≈25 μm (Figure S23c, Supporting Information) -comparable to the 25 μm PMBs and particles.To verify the hollow structure, the shell was stained with Nile red and CLSM clearly revealed the absence of a fluorescent core (Figure 5b).The sonication of PPGDA/DSDMA capsule suspensions showed no detectable increase of fluorescence intensity using the TSP method (Figure 5c), quantitatively leading to <1% productive bond scission and no visually identifiable broken fragments (Figure 5d).

Influence of the Shell Stiffness
In the following, we investigated the relationship between the activation performance of disulfide mechanophores and the mechanical properties of the PMB shells, in particular shell stiffness.Therefore, we prepared PMBs with variable relative disulfide crosslinker concentrations (v/v) at 5, 10, and 20% (Figure 6a).Since the DSDMA crosslinker had a significantly shorter contour length compared to the PPGDA crosslinker, we hypothesized that the average mesh size would decrease and the stiffness of the network would increase with increasing DSDMA concentration.AFM was carried out using an FMV-A tip with a resonance frequency of 75 kHz loading 15 nN force to assess the shell stiffness (Figure S24, Supporting Information) and verified this assumption. [25]While 5% DSDMA PMBs showed a stiffness of 0.28 ± 0.01 N m −1 (Figure 6b), those with 10% were determined to0.52 ± 0.06 N m −1 (Figure 6c), and those with 20% to 2.39 ± 0.33 N m −1 (Figure 6d).The increasing PMB stiffness with increasing DSDMA concentration (Figure 6e) was rationalized on the one hand by the decreasing average mesh size and hence lower elasticity and on the other hand by the observed increasing thickness of the PMB shells as visualized by SEM of open cracks after sonication (Figure S25a-c, Supporting Information).PMBs with 5% DSDMA exhibited the thinnest shell according to these measurements (Figure S25d).The disulfide bond scission capability of the PMBs was then investigated by ultrasonication.Compared to the 10% and 20% DSDMA variants, PMBs with 5% DSDMA showed the highest increase in fluorescence intensity after TSP treatment (Figure 6f-h) with >20% productive bond scission, which is eight times higher than the 20% DSDMA PMBs (Figure 6i).We hypothesized that the flexible polymeric network in combination with the low shell thickness synergistically contributed to the activation efficiency.

Computational Microbubble Simulation
To further rationalize the experimental results exploring the parameter space of PMB shell thickness and crosslink density, we investigated the deformation behavior of PMBs under US through physical calculations.The oscillation of the outer radial stretches ( = r/R 0 ; with r being the current outer radius upon US and R 0 being the initial outer radius) of the PMBs upon US occurred very rapidly (at the μs scale) owing to the sinusoidal wave of the driving pressure (Figure 7a). 25 μm PMBs showed the largest radial stretches compared to 35 and 50 μm PMBs.Simultaneously, the eigenfrequencies (f 0 ) of specific initial PMBs with diameters of 25, 35, and 50 μm PMBs were calculated to 21.09, 14.8, and 10.22 kHz (Figure 7b) through the equation where  is the polytropic index, p g0 is the initial gas pressure of the PMB, μ is the shear modulus of the PMB shell,  is the thickness of the PMB shell, and  is the density of a H 2 O/MeCN solution.All calculation procedures and involved parameters can be found in the Supporting Information.We found that the eigenfrequency of the PMBs decreased with increasing size.By using the same procedure, radial stretches of 25 μm PMBs with different DSDMA crosslink densities were additionally calculated.The largest radial stretches occurred on PMBs with 5% DSDMA crosslinking density (Figure 7c).Furthermore, as an investigation example, the radial stretch distribution of a 25 μm PMB shell with 5% DSDMA was simulated on a spatiotemporal level.The time-lapsed visual distribution of the PMB shell showed a suddenly increasing tendency of radial stretches ≈112 μs US application (Figure 7d).

Discussion and Conclusion
We demonstrated the successful application of microfluidic synthesis methods to produce a PMB system comprised of a gas core and a disulfide mechanophore-crosslinked polymer shell.We showed that productive mechanochemical disulfide bond scission and shell fragmentation of these structures was observed in the timeframe of 15 min using the US at a frequency of 20 kHz, which was considerably accelerated compared to the mechanochemical activation of disulfide mechanophorecentered-linear polymer chains.Thereby, we expanded the material space that relies on the alteration of polymer topology to steer mechanochemical bond scission where polymer brushes, [14][15][16][17] dendrimers, [17][18][19] micelles, [20] and microgels [21][22][23][24][25] have been investigated before.
Our experiments strongly suggested that the gas cores of the PMBs played a crucial role in this topological activity  enhancement by serving as cavitation nuclei thus leading to stronger inertial cavitation.This interpretation was supported by variation of the PMB diameters from initially 25 over 35 to 50 μm.It was reported before that variations in gas bubble size affected the inertial cavitation performance [43] and this was reflected in decreasing mechanochemical disulfide bond scission with increasing PMB diameter in our observations as well.Computational investigations using near-resonance excitation [45,46] supported our findings suggesting that the US-induced deformation of the PMB shells was largest when the eigenfrequency of the PMB approached the frequency of the used 20 kHz US source.Additional use of high frequency focused US at 0.68 MHz far off-resonance with regard to the three produced PMB diameters underlined this interpretation since no significant difference in mechanochemical reaction yield could be detected under these conditions.Nevertheless, it is plausible that in addition to US resonance, regular cavitation processes also contributed to bursting the PMBs.For example, Suslick and coworkers showed that microparticles collide with high speed upon 20 kHz US application. [47,48]Likely, this created a base level of mechanochemical activation even in off-resonance situations.
Moreover, we investigated alternative internal structures that were produced with comparable diameters for their mechanochemical performance as control materials.We neither observed notable mechanophore activation upon ultrasonication with solid core-containing polymer network particles nor with capsules filled with water instead of gas, although the polymer network composition and particle size was identical to the PMBs.These materials thus constituted an important control group ruling out effects stemming from either the chemical composition or size of the objects in relation to their interaction with ultrasound.
The investigation of increased disulfide-crosslinker content then helped us to understand the parameter space governing efficient mechanochemical activation in more detail.We hypothesized that the shorter disulfide crosslinker would decrease the average mesh size thus increasing the stiffness of the bubble shell and we confirmed this by AFM indentation measurements.In addition, we found that the shell thickness increased with increasing crosslink density.These experiments underlined the multiparameter character of the optimization of polymer systems for mechanochemistry.While very loosely 5% DSDMA crosslinked networks were fractured swiftly and led to ≈20% relative productive bond scission, increasing the DSDMA concentration to 20%, and thereby shell thickness, decreased the relative fraction of obtainable disulfide scission to ≈2.5%.Not all of the present disulfide mechanophores were activated by force, which was likely caused by the fragmentation of the polymer shell below the limiting molar mass required for the transmission of shear force to all available mechanoresponsive bonds.We additionally reasoned that on the one hand cavitation bubble nucleation was possibly negatively affected by the increasing thickness of the PMB shell.On the other hand, we made similar observations before on strongly crosslinked microgel cores where the lack of chain segment solvation and thereby the ability to respond to ultrasound-induced shear was found to be responsible. [22]Computational investigations then allowed us to decouple the DS-DMA crosslink concentration from the PMB shell thickness.We found an increased US-induced deformation of the PMB shells with lower DSDMA content and therefore lower stiffness, which rendered the mechanical deformation of stiff PMBs more difficult.
Overall, we found that thin shells, low shell stiffness and crosslink density, and a PMB size-dependent eigenfrequency close to the used US frequency were beneficial to maximize the mechanochemical yield over the course of the sonication process.
We believe that our proposed microfluidic fabrication method is in principle universally applicable to all mechanophores relying on covalent bond scission to generate a functional moiety.Yet, the reactant and experimental design have specific requirements that must be balanced carefully.Next to the microfluidic optimization of the associated phases in flow velocities and viscosities to obtain stable G/O/W emulsions, the polarities of the reactants and specifically the mechanophore-containing monomer must be considered so that they remain exclusively in the desired phase.Nevertheless, we showed that including micron-sized gas bubbles into polymer, structures is viable to decrease the necessary required ultrasound energy for efficient bond scission.In the future, this strategy might be useful for developments in the field of sonopharmacology, [1] yet further experiments using medical ultrasound must be performed to assess this opportunity.Since the PMB production using the microfluidic approach is at the lower edge of the available size space and diameter reduction to obtain higher resonance frequencies will be difficult, this constitutes a formidable challenge.
Thiol Detection After Disulfide Bond Cleavage: TSP (100 μL, 1 mm) was dissolved in DMSO and added to the sonicated mixture to detect the cleavage of disulfide bonds directly after sonication.After 3d (to assure equilibrium conditions), the supernatant solution was transferred into a quartz cuvette and the fluorescence intensity at 530 nm was measured using a spectrophotometer.
Quantification of Cleaved Disulfide Bonds by the TSP Method: Calibration curves for thiol concentration and fluorescence intensity correlation were prepared from mercaptoethanol solutions with 15.625, 31.25,62.5, 125, 250, 500, and 1000 μm; as well as 1000, 2000, 3000, 4000, and 5000 μm in DI H 2 O. Subsequently, 500 μL of MeCN, 300 μL DI H 2 O, 100 μL TSP (1 mm), and 100 μL as-prepared mercaptoethanol solution were mixed.Here, pure MeCN was not used since it caused the collapse of PMBs.After 3 d, the fluorescence intensity of the solution was measured.Two equations to describe the relationship between thiol group concentration (0-100 μm and 100-500 μm) and fluorescence intensity were processed in Origin2018 (Figures S18 and S19, Supporting Information).To quantify the cleaved percentage of disulfide bonds, the total number of disulfide bonds must be known.TCEP (50 μm) was used to chemically cleave all disulfide bonds within the PMBs, PPGDA/DSDMA particles, and PPGDA/DSDMA capsules (Scheme S4, Supporting Information).Subsequently, DI H 2 O was used to wash the PMBs, particles, and capsules twice.Then, the TSP was employed to detect and quantify the concentration of thiol groups.The fluorescence intensity of sonicated samples was measured using a spectrophotometer and further used to calculate concentrations.The number of disulfide bonds and cleaved disulfide bonds within PMBs was calculated according to the equation: N = n•N A (N: number of atoms; N A : Avogadro constant).Thus, the cleaved percentage of disulfide bonds was calculated by: D number of total thiol groups; N c : number of cleaved thiol groups; N 0 : correction factor caused by weak background fluorescence of TSP).
Sonication Experiments: Before sonication, the number concentration of PMBs (2.2 × 10 6 mL −1 ), PPGDA/DSDMA particles (3.2 × 10 6 mL −1 ), and PPGDA/DSDMA capsules (2.9 × 10 6 mL −1 ) was regulated using an EVE Automatic cell counter (NanoEntek).Then, 10 μL of the respective emulsion was added to a 1.5 mL EPP tube consisting of 500 μL MeCN and 390 μL deionized H 2 O, and then sonicated for 1, 3, 5, 10, and 15 min.Therefore, a 20 kHz Q125 Sonicator (Q-Sonica) with a transducer probe (part no.4422, probe tip diameter 3.2 mm) was used at 50% of the maximum amplitude of 180 μm in pulsed mode with 2 s on and 1 s off.The power intensity I P of applied ultrasound was 6.94 W cm −2 calculated by I P = P•A −1 , where P is the output power (0.486 W) obtained via the spent energy over the sonication on-time t (P = E•t −1 ) and A is the surface area of the transducer probe (A = •r 2 = 0.07 cm 2 ).
For high-frequency focused ultrasound experiments, samples were dosed into a 96-well plate lumox 94.6120.096(SARSTEDT AG & Co. KG, Germany) having a transparent for ultrasound foil base.The well plate was sealed with a gas-permeable adhesive film 4ti-0516/96 (Azenta Life Sciences).A 3D-manipulator positioned the well plate on top of deionized H 2 O in a 150 L test tank.A US transducer PA1470 (Precision Acoustics Ltd, UK) was powered from a linear amplifier AG1021 (T&C Power Conversion, Inc., USA) with an external waveform generator 33511B (Keysight).The alignment of the transducer and well plate was done using a 0.5 mm needle hydrophone NH0500 with preamplifier HP and coupler DCPS, whose signal was observed on an oscilloscope DSOX3024T (Keysight).The focal point of transducer was positioned in the center of the well, 3 mm above its bottom.Voltage applied to the transducer was measured by the oscilloscope and adjusted to obtain the desired acoustic power based on the transducer's factory calibration data extrapolated to low voltages.Two seconds long US bursts (0.68 MHz, 1.6 W) were applied with 1 s intervals for 5 min.For the transducer used, the −6 dB focal diameter was 4.49 mm.I P was estimated by dividing the power by the focal area to ≈10 W•cm −2 .

Figure 1 .
Figure 1.Formation and characterization of PMBs.a) Schematic of 5% DSDMA-crosslinked PMB preparation in a G/O/W double emulsion microfluidic device, photopolymerization with a UV LED, and subsequent activation of disulfide mechanophores upon US application.b) Formation of PMBs in the microfluidic device.Inner channel (white): N 2 stream; middle channels (orange): oil phase; outer channels (blue): aqueous phase.c) Collection of PMBs in a glass vial after polymerization.d) Optical micrograph of as-prepared PMBs.e) Diameter distribution histogram of PMBs (N = 100) was calculated from panel (d).f) SEM image of PMBs.Inset: PMB with an opening crack.The shell thickness is ≈670 nm.CLSM micrographs of a Nile red-stained PMB: g) bright-field, h) fluorescence, and i) overlay of panels (g) and (h).

Figure 2 .
Figure 2. The activation of disulfide mechanophores within PMBs and the method of thiol detection.a) Schematic of disulfide activation under US application.Time-lapse micrographs of PMBs after b) 0, c) 10, and d) 15 min pulsed US treatment (20 kHz, 2 s on, 1 s off).e) Mechanism for the detection of disulfide bond cleavage using a thiol-selective fluorogenic probe.f) Photographs of PMB suspensions before (left) and after (right) US application under UV illumination.

Figure 3 .
Figure 3. Quantification of disulfide mechanophore activation within PMBs using TSP after US application.a) Relationship between thiol concentration and fluorescence intensity at 530 nm.b,e,h) Fluorescence spectra of PMB suspensions of 25, 35, and 50 μm PMBs after US treatment.c,f,i) Quantification of disulfide bond cleavage in 25, 35, and 50 μm PMBs after US treatment.d,g) Micrographs of PMB production in a microfluidic device for 35 and 50 μm PMBs.Mean values ± SD from the mean.N = 3 independent sonication experiments.

Figure 4 .
Figure 4. US-activated disulfide mechanophores in solid core PPGDA/DSDMA particles.a) Schematic of microfluidic particle formation in an oil-inwater (O/W) emulsion.b) CLSM micrographs of solid particles stained with Nile red.c) Fluorescence spectra of particle suspension over sonication time after TSP treatment.d) Quantification of disulfide bond cleavage.Insets are the micrographs for 0 and 15 min US treatment.Scale bars: 100 μm.Mean values ± SD from the mean.N = 3 independent sonication experiments.

Figure 5 .
Figure 5. US-activated disulfide mechanophores in PPGDA/DSDMA microcapsules.a) Schematic of microfluidic capsule formation in a water-in-oil-inwater (W/O/W) emulsion.b) CLSM micrographs of capsules stained with Nile red.c) Fluorescence spectrum of capsule suspension over sonication after TSP treatment.d) Quantification of disulfide bond cleavage.Insets are the micrographs at 0 and 15 min US treatment.Scale bars: 100 μm.Mean values ± SD from the mean.N = 3 independent sonication experiments.

Figure 6 .
Figure 6.Relationship investigation between activation performance of disulfide mechanophores and stiffness and shell thickness of PMBs upon US treatment.a) Schematic of PMBs with different DSDMA concentrations for the measurement of their shell stiffness by AFM.AFM stiffness measurements of PMBs with b) 5% DSDMA, c) 10% DSDMA, and d) 20% DSDMA.Insets are the corresponding SEM micrographs.All scale bars: 20 μm.e) Stiffness histogram.Mean values ± SD from the mean.N > 60 indentations per loading force.Fluorescence spectra of PMB suspensions with f) 5%, g) 10%, and h) 20% DSDMA.i) Quantification of disulfide bond cleavage with progressing sonication time.Mean values ± SD from the mean.N = 3 independent sonication experiments.

Figure 7 .
Figure 7. Theoretical calculation of PMB deformation upon ultrasonication.a) Outer radial stretches of PMBs at different initial outer diameters.The maximum stretch of ≈2.43 was observed at an initial diameter of 25 μm.b) Eigenfrequency of 25, 35, and 50 μm PMBs.c) Outer radial stretches of PMBs with DSDMA crosslink densities.The maximum stretch of ≈2.36 was observed for PMBs with 5% DSDMA shell and a shell thickness of 0.26 μm.d) Radial stretch distributions of 25 μm PMB with 5% DSDMA shell and a shell thickness of 0.26 μm at 0, 54.24, and 112.23 μs ultrasonication.