Interfacial Force‐Focusing Effect in Mechanophore‐Linked Nanocomposites

Abstract Enhanced force transmission to mechanophores is demonstrated in polymer nanocomposite materials. Spiropyran (SP) mechanophores that change color and fluorescence under mechanical stimuli are functionalized at the interface between SiO2 nanoparticles and polymers. Successful mechanical activation of SP at the interface is confirmed in both solution and solid states. Compared with SP‐linked in bulk polymers, interfacial activation induces greater conversion of SP to its colored merocyanine form and also significantly decreases the activation threshold under tension. Experimental observations are supported by finite element simulation of the interfacial stress state. The interfacial force‐focusing strategy opens a new way to control the reactivity of mechanophores and also potentially indicates interfacial damage in composite materials.


Materials
Unless otherwise states, all reagents were purchased from commercial source and used as received.
Deuterated solvents (chloroform-d, dimethyl sulfoxide-d6) were purchased from Cambridge Isotope Laboratories, Inc. Methyl acrylate (MA) was passed through a basic alumina filled column to remove inhibitors and subsequent bubbling with nitrogen to eliminate any oxygen. Copper wire was purchased from Fisher scientific. Micron-sized SiO2 particles were purchased from Polysciences, Inc. SiO2 nanoparticles were donated by Nissan Chemical in a dispersion of SiO2 nanoparticles with a size of 10-15 nm in methyl isobutyl ketone (30-31 wt% SiO2)

Characterization
Column chromatography was performed on a Biotage Isolera system using SiliCycle SiliaSep HP flash cartridges. NMR spectra were recorded using a Varian 500 MHz spectrometer or Carver-Bruker 500 spectrometer. Spectra were referenced to the residual solvent peak: Dimethyl sulfoxide Digital images of tensile specimens were taken using a digital camera (Canon G16, 12.1 megapixels). Fluorescence images were acquired with a confocal microscope (Leica TCS SP8).
The surface morphologies of micron-sized SiO2 particles were observed with an environmental scanning electron microscope (ESEM, FEI Quanta FEG 450). Transmission electron microscope (TEM) micrographs were taken using a JEOL 2100 cryo-TEM at an accelerating voltage of 120 kV. For examining the cross-sectional morphology of composites, the specimen was sectioned using a Leica microtome Ultracut UCT and placed on copper grids.

Instability of SP in ammonia solution
12 (5 mg, 0.013 mmol) was dissolved in 2ml of THF or toluene, followed by adding 0.1 ml of base.
To confirm the photochromic behavior, UV light was exposed to the solution. After adding ammonia solution (30 wt% in water), SP loses photochromic behavior, indicating the degradation of SP structures. Base-catalyzed hydrolysis is the possible degradation mechanism that generates Fisher's base and salicylaldehyde from the merocyanine form. [3] With the addition of triethylamine, SP retains the reversible photochromic behavior in both polar (THF) and non-polar (toluene) solvent. Therefore, we chose triethylamine as a basic catalyst for surface functionalization of SiO2 particles.

Surface-initiated SET-LRP from SP functionalized SiO2 particles
In a typical reaction, a Schlenk flask was charged with 50 mg of bromine-functionalized SiO2 particle and 4 cm of copper(0) wire under N2 atmosphere. 1 ml of DMSO was added to the flask and additional sonication was performed to disperse the SiO2. Then, 1 ml of methyl acrylate and 16 µL of Me6TREN were injected, followed by three cycles of freeze-pump-thaw. After the flask was filled with N2, the solution was stirred in a water bath for 2 h. The viscous solution was diluted with small amount of THF and precipitated in cold MeOH to yield polymer-grafted SiO2 particles.
The polymer composites were further dried in a vacuum oven at 80 °C overnight to remove residual solvents. Control samples were prepared in the same manner.

Surface-initiated polymerization from micron-sized particles
Following the general method described in Section 4.1, PMA was successfully grown from the surface of SiO2, confirmed by SEM ( Figure S3). For comparison to the polymer-grown particles, TEM images of bare SiO2 particles are included ( Figure S4). Figure S4. TEM images of SiO2 nanoparticles. a, Bare SiO2 nanoparticles (diameter = 10-15nm). b, PMA grown particles. Inset: Enlarged view of the SiO2 nanoparticles. In contrast to PMA-grown SiO2 nanoparticles, their surface was clean, and no polymer residue was observed.

Calculation of grafting density
Following the method described by Li et al., [4] the grafting density of PMA on the SiO2 particles was characterized. The weight percent of grafted PMA (x) and SiO2 (y) were determined using thermogravimetric analysis (Figure S5a, b). Also, the molecular weight and dispersity of grafted PMA was characterized by GPC after removing SiO2 with HF treatment (Figure S5c). can increase the Tg of the polymer. [5,6] The ester group from PMA is believed to form hydrogen bonds with hydroxyl groups on the SiO2, [7,8] so we assume that the Tg of the attached PMA on the SiO2 is higher than that of free PMA if both of them have the same molecular weight. We measured the Tg of attached PMA to the SiO2 and free PMA obtained by removing SiO2 with HF by varying the molecular weight of PMA (Figure S6Error! Reference source not found.). At lower molecular weights, the attached polymers on the SiO2 have a higher Tg than free polymers, while this difference is reduced as the molecular weight increases. This difference might be related with the different segmental mobility of polymer depending on the distance from the attractive substrates. [9] Since low molecular weight PMA can form a thinner layer on the SiO2, the chain mobility is greatly reduced by the attractive interaction from the surface. This effect is minimized when the PMA forms a thicker layer. The Tg dependence on molecular weight was fitted to the Flory-Fox equation. (1) where Tg ∞ is the maximum Tg at a theoretical infinite molecular weight and K is polymer specific constant related with the chain-end free volume. In both cases, Tg ∞ is in the similar range of 20 °C, while K value decreases in half for attached PMA indicating the reduced free volume by attractive interaction with SiO2.

Tension test
Dog-bone specimens were tested was uniaxially deformed at a strain rate of 0.1 s -1 by two opposing actuators while capturing optical images. Load was recorded using a 50-lb capacity load cell (Honeywell Sensotec) attached to one of the actuators. The acquired load-displacement data were converted to engineering stress and strain. Fluorescence intensity was calculated by averaging the red channel intensity of the gauge section of the sample. The stress-deformation ratio curve for each sample and the optical images of tensile samples are summarized in Figure S8. Only active sample exhibited color change in the gauge area under tension ( Figure S9). Figure S9. Tension test of PMA composites containing active particles. Dog-bone shaped specimen change color in the gauge area while stretching. Dog-bone specimens were tested with a combined mechanical and optical testing setup. [10] A 532nm laser (0.6 mW) was used for the excitation light source for fluorescence measurements. The specimen was uniaxially deformed at a strain rate of 0.1 s -1 by two opposing actuators. Load was recorded using a 50-lb capacity load cell (Honeywell Sensotec) attached to one of the actuators.

Characterization of attached SP to SiO2 particles
The attached amount of SP to SiO2 was determined by TGA, which was around 2.5 wt% ( Figure   S14). were added to the vial and sonicated. The final mixture was injected the glass molds while flushing with N2 in a zip-lock bag. After at least 12 h, the samples were removed from the mold and washed with methanol to remove residual monomers. The final films were dried in a vacuum oven and laser cut into dog-bone shaped specimens.
Four or five of each specimen type were tested and the stress-deformation ratios were recorded ( Figure S15).
A generic area was selected from TEM images of the cross-section of the xPMA/SiO2 composite and then modeled as shown in Figure 4a of the manuscript. In particular, the analyzed area contains eight silica particles with a diameter of 18 nm, embedded into a PMA matrix, and located at different distances between each other.
The PMA/SiO2 composite was modeled as a 2D deformable body. While SiO2 was considered a simple elastic material (Young's modulus= 73 GPa, and Poisson's ratio= 0.15), hyperelastic properties were assigned to the PMA matrix. Specifically, the Arruda-Boyce hyperelastic model was used to describe the strain energy potential, and the model coefficients were obtained by fitting experimental test data of PMA samples. [11] A linear displacement was applied to the two free edges along the y direction (see Figure   4) until reaching a final deformation ratio of 3.5. The 2D model was discretized by CPS8R elements, and mesh refinement studies were performed in order to validate the accuracy of the model.