Hierarchically Structured Deformation‐Sensing Mechanochromic Pigments

Abstract Mechanochromic materials alter their color in response to mechanical force and are useful for both fundamental studies and practical applications. Several approaches are used to render polymers mechanochromic, but they generally suffer from limitations in sensing range, capacity to provide quantitative information, and their capability to enable broad and simple implementation. Here, is it reported that these problems can be overcome by combining photonic structures, which alter their reflection upon deformation, with covalent mechanophores, whose spectral properties change upon mechanically induced bond scission, in hierarchically structured mechanochromic pigments. This is achieved by synthesizing microspheres consisting of an elastic polymer with spiropyran‐based cross‐links and non‐close‐packed silica nanoparticles. A strain of less than 1% can be detected in a shift of the reflection band from the photonic structure, while the onset strain for the conversion of the spiropyran into fluorescent merocyanine ranges from 30% to 70%, creating a broad strain detection range. The two responses are tailorable and synergistic, permitting the activation strain for the mechanophore response to be tuned. The mechano‐sensing photonic pigments are demonstrated to be readily incorporated into different polymeric materials of interest and quantitatively probe spatially heterogeneous deformations over a large strain range.


4
Scientific Instruments) to prevent charging. Electron micrographs obtained from the FIB-SEM were corrected for the tilt of the stage (52º).
Photographs. Photographs and videos of samples were recorded with a Nikon D7100 digital camera equipped with an AF-S DX Zoom-NIKKOR 18-135mm lens (f/3.5-5.6G IF-ED) in manual focus mode under ambient light.
Fluorescence microscopy. Fluorescence microscopy images were acquired at 5x or 20x magnification using an Olympus BX51 microscope equipped with an Olympus DP72 high-resolution camera. The samples were imaged in reflectance mode using an X-Cite Series 120-Q Mercury vapor short arc lamp as the excitation source (brightfield and λex = 560 nm). A standard white diffuser was used as a white reference sample.
Microscopic reflectance spectroscopy. Microspectroscopy was performed with a Zeiss Axio Scope.A1 polarized light microscope, connected to a Point Grey GS3-U3-28S5C-C CCD camera (FLIR Integrated Imaging Solutions Inc.). For all measurements, the samples were imaged in reflection mode (bright field) with Koehler illumination, using a halogen light source (Zeiss HAL100). Dry pigments, or pigment-containing PDMS or LLDPE were deposited on a glass slide and imaged in reflection in bright field. To measure the spectral reflectance of the pigments, an optical fiber (Ocean Optics, 230 µm core size) was positioned confocal to the image plane, with the other end of the fiber coupled to a diode-array spectrometer (FLAME-T-XR1-ES, Ocean Insight). Normalization of the reflectance spectra was performed against a silver mirror (Thorlabs, PF10-03-P01, avg. reflectance >97.5% for λ = 450-2000 nm). The spectra were collected using a 20x objective (Zeiss EC Epiplan-Apochromat, NA = 0.6).

Compression of individual pigment particles.
Deformation was applied at a strain rate of 0.25-0.5 % s-1 (indenter velocity 0.5 μm s-1) using a motorized actuator (MTS50E/M, Thorlabs with bidirectional repeatability ±0.8 µm) controlled via a computer interface (Kinesis software). Load data were obtained from a load cell (LSB205, maximum load capacity 8.90 N, FUTEK). Images of the pigment prior to and under compression were acquired with a microscope (further details on the in-situ optical characterization are provided below). The set-up was installed on an anti-vibration table. Individual pigments were selected manually with a pipette or the tip of a micro-spatula and deposited onto a glass plate with a thickness of 3 mm. The compression tester consisted of a piece of Si (100) wafer adhered to a small screw, which was mounted in the load cell. The initial pigment diameter was obtained from micrographs taken prior to compression. To record the reflectance spectra in situ, the pigments were continuously compressed from above. The indentation experiments were initiated with the indenter slightly above the top of the pigment to allow the indenter to reach a steady-state velocity before touching the pigment, and terminated at 80% compressive axial strain. To record confocal images in situ, the compressive motion was paused at a given strain to record the images.
In situ reflectance microscopy. Microspectroscopy during compression experiments was performed using a custom-built inverted light microscope and a CMOS camera (Blackfly S BFS-U3-200S6C-C, 5 FLIR Integrated Imaging Solutions Inc.). A high-power Xenon light source (HPX-2000, Ocean Optics) was used as a light source for all measurements. Dry pigments were deposited on a thick glass slide (3 mm thickness) and positioned below the indenter; the pigments were imaged from below the glass slide in reflection in bright field. The spectral data were collected in reflection mode (bright field) using a 50x objective (BD Plan Apo, Mitutoyo, NA = 0.55). An optical fiber (Ocean Optics, 230 µm core size) was positioned confocal to the image plane, with the other end of the fiber coupled to a diode-array spectrometer (FLAME-T-XR1-ES, Ocean Insight) in order to measure the spectral reflectance of the pigments. Normalization of the reflectance spectra was performed against a silver mirror (Thorlabs, PF10-03-P01, avg. reflectance >97.5% for λ = 450-2000 nm). The spectra were recorded continuously (1 spectrum/s) for the duration of the experiment, without refocusing. Spectral changes were similar when the pigments were observed upon stopping the indenter and subsequent refocusing.
In situ confocal microscopy. Images were acquired with an inverted Nikon Eclipse TE2000-U microscope using an excitation laser of 543 nm (Melles Griot) and a 10x objective (Plan, Nikon, NA = 0.25). For each compression step, the imaging plane was refocused manually to ensure that the edges of the pigment were in focus at its largest diameter (confocal slice thickness ~30 µm). Brightfield micrographs were acquired with a digital camera (Evolution MP, Media Cybernetics).
Tensile testing. Uniaxial tensile tests were carried out with rectangular samples with dimensions of 4 × 0.3 mm (width × thickness) that were cut from compression-molded films. The distance between the clamps at zero strain, which defined the initial length of the sample, was 18 mm. A pre-load of approximately 0.1 N was applied before starting the tensile tests. Stress-strain data collected during insitu microscopic optical and fluorescence imaging were recorded at a strain rate of 0.56-0.57% s -1 (velocity of clamps 0.1 mm s -1 ) using a Linkam TST350 microtensile stage that was equipped with a 20 N load cell and controlled by accompanying Linksys32 software. The experiments were conducted under ambient conditions (T = 23-24 °C).
Indentation of PDMS films containing mechano-pigments. The same set-up and experimental parameters were used for the indentation experiments as for the compression experiments described above, except the Si (100) wafer was replaced with a steel sphere (diameter 5 mm). The PDMS film (thickness 0.8 mm) was prepared on a glass slide, mounted on the indentation stage and indented from below, while being imaged from above with a camera (Nikon D7100) to record the photonic color changes in the pigments. To record the spiropyran activation, the film was indented, then imaged post mortem with a fluorescence microscope (Olympus BX51).

Methods and Procedures
Synthesis of silica particles. An aqueous dispersion of seed nanoparticles was first prepared by sonicating silica seed nanoparticles (5 g) in ethanol (100 mL) in a 250 mL beaker, first in a bath sonicator 6 (Huber) for 1 h, then at 40-50% maximum amplitude with a horn sonicator for 2 h (Branson Digital Sonifier, 1 s on, 2 s off). 10 mL of the aqueous silica seed dispersion, ethanol (150 mL), water (35 mL) and NH 4 OH (30% aqueous solution, 5 mL) were added to a 500 mL round bottom flask, with a magnetic stir bar. TEOS was added via a syringe pump at 0.8 mL/h, stirring at 400 rpm. The reaction was left to stir for 18 h at room temperature. The dispersion was centrifuged, the supernatant replaced with fresh ethanol and the particles were redispersed in a bath sonicator; this process was then repeated twice. The particles were then dried for 18 h at 70 °C and redispersed in ethanol at a concentration of 50 mg/ mL.
The diameters and polydispersity of the silica particles thus made were analyzed from SEM images (Supporting Figure 2 and 3, Table 2).
Preparation of mechano-pigments. Three photonic compositions were studied, with the following parameters ( Table 1). The standard matrix composition for each photonic composition involved a crosslink density of 1 mol%, including a spiropyran content of 0.25 mol%. For photonic composition 2, the volume fraction of silica and cross-link density of the matrix was also varied (main article, Figure 2  To produce 0.5 mL of a silica/PEGPEA mixture, containing silica nanoparticles with a diameter = 163 nm, a silica volume fraction (φ(SiO2)) of 0.35, a SP cross-linker content of 0.25 mol%, and a total crosslink density of 1 mol%, a suspension of the silica nanoparticles in ethanol with a concentration of 57 mg/mL was prepared. The density of the SiO2 nanoparticles was assumed to be 2 g/cm 3 . [3][4][5]  Spectral and image analysis. Spectra, images and videos were processed and analyzed in ImageJ (Fiji) and MATLAB (2019b). In particular, the following MATLAB scripts were used to analyze specific microstructural and mechanochromic changes. To generate the radial distribution functions and histograms in Figure 2, the SEM images were corrected for the angle of the FIB-SEM stage (52º) and 8 the inter-particle distances were obtained with the imfindcircles algorithm, which uses a circular Hough transform. For the analysis of the reflectance spectra in Figure 2 in the main manuscript, a linear baseline was fitted to the spectra and subtracted from the intensity signal; a Gaussian was then fitted to each baseline-corrected spectrum, and the wavelength of maximum intensity was taken from the fitted curve and plotted against the strain of the pigment. Similar results were obtained using the center of gravity of the baseline-corrected measured spectra. The confocal images showing the mechano-activated fluorescence intensity in single pigments, depicted in Figure  Synthesis and characterization of spiropyran dimethacrylate cross-linker. Literature synthetic procedures were followed [1,2] and the product was characterized by 1 H NMR spectroscopy.  10 Figure S3. Size analysis of the colloidal silica particles shown in the SEM images in Figure S2. Table S2. Size analysis of the colloidal silica particles shown in the SEM images in Figure S2. Errors in the silica particle diameter are standard errors in the mean, calculated from the analysis on three separate SEM images.   12 Table S3. Expected and measured volume fractions of silica, φ(SiO2) in three pigment batches.
The surface layer of silica particles in the mechano-pigments seems to be close-packed ( Figure S8-S10), which possibly results from diffusion of the PEGPEA monomer into the continuous aqueous phase in the emulsion step of the synthesis. The spectral measurements ( Figure S5 and S12) indicate that the optical properties are determined by the bulk, rather than the surface layer.   The predicted distance between nearest neighbors, a, defined as the distance between the centroids of two neighboring particles, was calculated assuming a face-centered cubic close-packing: [6] = ( π 3√2 (SiO 2 ) ) 1 3 (1) where d is the particle diameter and (SiO 2 ) is the volume fraction of particles.
The predicted wavelength of the light most strongly reflected by the particle arrangement was calculated using Equation 2: where a is the interparticle spacing and eff is the average refractive index, which is calculated as where (SiO 2 ) and (PEGPEA) are the refractive indices of the silica particles and the polymer matrix of the pigment, respectively.    Text S3: Calculation of mechano-activated fluorescence intensity.
The normalized fluorescence intensity from the merocyanine was calculated to enable a comparison between the mechano-activation in tension and in compression. Normalized fluorescence, normalised ( ) was defined as: [7,8]      and in tension (incorporated in LLDPE) vs. strain along the axis of the applied strain. In the case of compression, the periodicity decreases and a blue-shift is observed to occur along the axis of compression, whereas in the case of tension, the blue-shift is observed along the axis perpendicular to the axis of applied tensile strain.

Text S4: Indentation experiments on mechano-pigments incorporated in PDMS.
For indentation experiments, the PDMS film containing 10wt% mechano-pigments was left attached to the glass slide. A film of carbon black-containing PDMS was positioned below the film containing mechano-pigments, and the steel letter stamp 'a' (Figure 23a) was lightly pressed by hand into the mechano-pigment film from below, while taking a photograph from above (Figure 16b). On applying further pressure by hand, the mechano-activation of spiropyran to purple merocyanine could be observed following removal of the stamp (Figure 23c). The spectral changes upon indentation were also recorded with a spectral microscope (Figure 24). Similar colorimetric changes could also be observed upon fracturing or damaging the film with a spatula (Figure 25), and the purple mechano-activated merocyanine could furthermore by transformed back to colorless spiropyran upon exposure to green light ( Figure 25c). Pigments not containing spiropyran did not show purple coloration upon indentation ( Figure 26).    Figure S28. Force vs. indentation depth for PDMS containing 10 wt% mechano-pigments, where indentation depth is defined as indentation distance/ initial film thickness.