Mechanoresponsive Self‐Assembled Perylene Bisimide Films

Abstract In this work, self‐assembled amino‐acid appended perylene bisimides (PBIs) have been studied that when processed into thin films change their resistivity in response to being bent. The PBIs assemble into structures in water and form thin films upon drying. These normally delicate thin films can be tolerant to bending, depending on the aggregates they form. Furthermore, the films then reversibly change their resistivity in response to this mechanical stimulus. This change is proportional to the degree of bending of the film giving them the potential to be used quantitatively to measure mechanical movement, such as in wearable devices.


Film Preparation.
Thin films were prepared on a Lyreco PVC clear covers that were cut to the required size. 20 μL of the required solution was dropped using a pipette inside a 5 x 5 mm sticky tape mask and leaving to air dry. Once dried, the mask was removed to produce a 5 x 5 mm square. Two silver electrodes were added to the edge of the sample using silver paste (Agar Scientific, Acheson Silver Electrodag 1415), ensuring that full contact was made to opposite sides of the sample square ( Figure  S1). Two pieces of copper tape (RS components) were then stuck to the silver electrodes so that they could be attached to the potentiostat. pH Measurements. pH measurements were performed using a FC200 pH probe (HANNA Instruments) with a 6 mm x 10 mm conical tip. The accuracy of the pH measurements is quoted as ± 0.1. 1 H, 2 H and 23 Na NMR. The experiments are based on previous work, where we have shown that 2NapFF and PBI-A align in a magnetic field, which leads to the residual quadrupolar coupling effects seen here. 2,3 This alignment occurs in a 5 mm NMR tube once inserted into the spectrometer and is field dependent. NMR spectra were acquired using a Bruker 400 MHz ( 1 H) spectrometer equipped with a broad band probe operating at 105.84 MHz for 23 Na. 23 Na spectra were acquired in 256 scans using a 12 µs 90-degree pulse, a signal acquisition time of 0.3 s, a relaxation delay of 0.1 s and a sweep width of 120 ppm. 2 H spectra were recorded via the lock channel with 128 scans, a 200 µs 90-degree pulse, signal acquisition time of 2.0 s, a relaxation delay of 0.2 s and a sweep width of 10 ppm. 1 H spectra were acquired using the perfect echo WATERGATE sequence of Adams et al. 5 incorporating the double echo W5 sequence of Liu et al. 6 The acquisition time was set at 2.5 s. The delay between successive hard pulses of the selective pulse train was set at 333 µs corresponding to a 3000 Hz spacing between the null points. The relaxation delay was set at 1 s. 4 dummy scans and 16 scans were acquired. 1 mL of samples of each sample was used for these experiments and prepared as described above with 0.1 vol% D2O added as a probe for 2 H NMR while 0.003 vol% THF was added to provide an internal reference signal. All spectra were calibrated to 1 H THF at 1.88 ppm. 7 Spectra were processed in Bruker Topspin 3.6.2. 23 Na spectra were processed with an exponential line broadening factor (LB) of 3 Hz. The linewidths at half height of 23 Na (Dn1/2) were obtained by Lorentzian deconvolution. T2 relaxation times of 23 Na were obtained using Equation S1: 8 (Eq.S1) The contribution to the linewidth from the magnetic field inhomogeneity can be neglected as analysis of the THF resonances confirmed it to be less than 2 Hz for 1 H ( Figure S4b).

UV-vis Absorption Spectroscopy.
Solid UV-Vis absorption data was obtained with an Agilent Cary 60 UV-vis spectrophotometer. Samples were cast onto the plastic used for the photoconductivity measurements. Data was collected between 350 -800 nm at 200 nm/min.

Optical Microscopy.
Optical microscope images were collected using a Nikon Eclipse LV100 microscope with a Nikon Plan ELWD 50x/0.60 or x5 lens attached to an Infinity2-1C camera.
Shear Induced Alignment. Shear alignment data was collected on an Anton Paar Physica 302 rheometer fitted with the RheoOptics accessory with cross polarisers. Camera Lumenera Lu165c: 12-bit, 1392 x 1040 pixels, 2/3" CCD-Sensor, USB 2.0 with an Edmund Optics Worldwide white Mi-LED fibre optic LED illuminator with iris light source.
Alignment was induced using a 25 mm parallel plate at 1000 s -1 at a gap of 0.1 mm. Samples were pipetted onto the plate (0.5 mL) and the plate lowered onto the solution.
The sample was trimmed before the measurement was started and checked for any bubbles on the camera. If any bubbles were present the solution was replaced. The tests were only started when there were no bubbles present, the presence of bubbles interferes with the images. Samples were allowed to shear for 5 minutes, but alignment if it happened was almost instantaneous and so images were taken using the Luminera software after 1 minute. When the shear was stopped, any sample that showed alignment quickly de-aligned as we have previously reported.
Atomic Force Microscopy. Thin films were prepared were prepared as described above, either on plastic or glass for testing. The films were studied using a Bruker Innova atomic force microscope. Each image consists of 512 lines with 512 points per line. The scan rate was 1.0 Hz for the experiments and the scan area of the plots is 20 × 20 μm unless otherwise stated. The images were analysed using the NanoScope Analysis 1.5 program by Bruker. The default bow remove processing for the software was carried out to level the plots. 9 Hole analysis was carried out using WsXM 5.0 Develop 9.4 software with holes defined as having a minimum area of 0.0061 μm 2 at below 50% of the maximum height.
Photoconductivity Measurements. Photoconductivity measurements were performed using a Palmsens Potentiostat using a two-electrode measurement. Linear sweep measurements were recorded from -4 V to 4 V at a scan rate of 0.05 V/s. A 'dark' base line reading was collected before any irradiation happened. Samples were irradiated at 1 cm away from the sample 365 nm LED powered with an ISO-TECH IPS303DD DC power supply, which operated at a constant current of 0.7 A for 5 minutes, or until the amount of radical stopped increasing. All measurements and samples were collected at least in triplicate to ensure reproducibly of the results.
Bending experiments. OpenSCAD was used to design the film holders which consisted of two pieces of plastic that fit together on top of the films on plastics as described above. The top piece of the holder had a square cut out to allow the film to be irradiated. The curvature radius was changed to get holders ( Figure S2 a and b) with a different curvature allowing the films to be bent to various degrees. The degree of bending was calculated using Equation S2. Five different angles were tested, 0.0°, 9.5°, 11.5°, 14.3° and 19.1°. (Eq.S2)

= ℎ 360°
These were exported to .STL format in order to print using a 3D printer ( Figure S2c). The holders were fixed together with screw to hold the film in place during the measurements and allowing the films to be removed without damage ( Figure S3). At least 9 repeats were carried out for each of the materials tested, however for the ease of reading of the document we have included 3 repeats, as they all showed the same behaviour, but had slightly different resistivities to start with, as they were all different films and we have found that there is some variability in the film resistivity to start with. The samples were irradiated for at least 10 minutes before measurements were carried out and were kept under constant 365 nm light to ensure they stayed conductive. The films were bent in order of increasing curvature and then were left to completely recover in the dark then re-irradiated and the data repeated to check recoverability. Randomly selected angles were then checked to ensure reliability of the results. The dashed red lines on (b) are Lorentzian fits to the spectra. The peak marked * has a fitted linewidth of 2 Hz in all samples, confirming that the 23 Na linewidths will be unaffected (< 0.5 Hz) by the magnetic field inhomogeneity and that THF does not interact strongly with the PBI structures. The broadness of the PBI resonances relative to THF indicates that the PBI are aggregated.