Direct Ink Writing of Recyclable Supramolecular Soft Actuators

Direct ink writing (DIW) of liquid crystal elastomers (LCEs) has rapidly paved its way into the field of soft actuators and other stimuli-responsive devices. However, currently used LCE systems for DIW require postprinting (photo)polymerization, thereby forming a covalent network, making the process time-consuming and the material nonrecyclable. In this work, a DIW approach is developed for printing a supramolecular poly(thio)urethane LCE to overcome these drawbacks of permanent cross-linking. The thermo-reversible nature of the supramolecular cross-links enables the interplay between melt-processable behavior required for extrusion and formation of the network to fix the alignment. After printing, the actuators demonstrated a reversible contraction of 12.7% or bending and curling motions when printed on a passive substrate. The thermoplastic ink enables recyclability, as shown by cutting and printing the actuators five times. However, the actuation performance diminishes. This work highlights the potential of supramolecular LCE inks for DIW soft circular actuators and other devices.


Synthetic procedure
The supramolecular cross-linked thermoplastic LCE is obtained by sequential thiol-acrylate and thiol-isocyanate addition reactions ( Figure  S1, Table S1). In short, reactive mesogens 1 (15.84 mmol, 9.312 g) and 2 (15.84 mmol, 10.6597 g) were added in a 1:1 molar ratio to a 250 mL round-bottom vessel, after which they were dissolved in DMAc 50 wt% (42 mL) at 50 °C under Ar atmosphere. Next, the temperature is brought back to room temperature, and an excess of 3 (38.00 mmol, 6.9394 g) and 0.1 wt% of 4 (0.28 mmol, 0.0381 g) are added. After allowing the reaction to proceed for two hours, the prepolymer is sequentially prepared via a base-catalyzed thiol-isocyanate addition reaction. For this reaction, an aliphatic diisocyanate 5 (12.70 mmol, 2.1330 g) in DMAc 50 wt% (2 mL) and an amine catalyst 6 (0.1 wt%, 0.34 mmol, 0.0346 g) is added to the thiol-terminated oligomer. After 15 minutes at room temperature, the viscosity increased, and therefore, additional DMAc (40 mL) was added to the mixture. Next, a 1:1 molar ratio of the dithiol chain extender 7 (6.34 mmol, 0.9533 g) is added to the oligomer prepolymer, after which the reaction proceeds overnight at 60 °C. The final polymer is obtained through precipitation in cold Et2O (2.5 L) overnight and subsequently stirred in fresh cold Et2O (400 mL) for additional 3.5 hours, yielding the PTU LCE (96.2% recovery). Finally, the material was dried at 40 °C under vacuum overnight.
Characterization FTIR spectra were recorded on a Varian 670 IR spectrometer equipped with an attenuated total reflectance (ATR) sampling accessory using a diamond crystal from 4000 to 650 cm -1 , collecting 50 scans per spectrum with a spectral resolution of 4 cm -1 . The FTIR spectra are recorded at room temperature and are processed with Varian Resolutions. Gel permeation chromatography (GPC) was performed on a Water HPLC system equipped with a PSS PFG (8 × 50 mm 2 , 7 µm) and two PFG linear XL columns (8 × 300 mm 2 , 7 µm) in series. The mobile phase, 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) with potassium trifluoroacetate (20 mM) at 35 °C, is supplied at a flow rate of 0.8 mL min -1 . The samples were prepared at room temperature in HFIP with potassium trifluoroacetate (20 mM) and toluene (20 mM). The molecular weights were determined using a refractive index detector relative to poly(methyl methacrylate) standards. Differential scanning calorimetry (DSC) thermographs are measured with a TA Instruments Q1000 DSC instrument, using hermetic T-zero aluminum sample pans. All scans were conducted with 10 ± 1 mg polymer over the temperature range from -50 to 210 °C with heating and cooling rates of 10 °C min -1 under N2 atmosphere. The transition temperatures are determined from the second heating cycle. TGA was carried out on a TA instruments Q50 instrument with 5.1 mg polymer over a temperature range from 28 to 800 °C and a heating rate of 10 °C min -1 . For the thermal degradation test with TGA, 7.6 mg polymer was heated to 200 °C for 16 hours. Dynamic mechanical analysis (DMA) was performed with a TA Instruments Q800 apparatus in vertical tension mode on samples of (8.8 × 5.7 × 0.4 mm 3 ) cut from compression-molded films. The thermographs were measured between -50 and 250 °C, with a heating rate of 5 °C min -1 , a single 1 Hz oscillating frequency, 20 µm amplitude, and 0.01 N preload force. X-ray scattering measurements were performed on a Ganesha lab instrument equipped with a Genix-Cu ultralow divergence source. This source generates X-ray photons with a wavelength and flux of 0.154 nm and 1 × 10 8 photons s -1 , respectively. Diffraction patterns were obtained using a Pilatus 300 K silicon pixel detector with 487 × 619 pixels of 172 × 172 µm 2 . As a calibration standard, silver behenate was used. The sample-to-detector distance is 89 mm for wide-angle (WAXS) configurations, whereas for medium angle (MAXS) configurations a distance of 440 mm is used. After data collection, a custom Python script with the PyFAI software package S2 is used to reduce and analyze the obtained data. The orientational order parameters were calculated from the diffraction patterns using the Kratky method. 1

Preparation of Glass Printing Substrates
Free-standing actuators were fabricated by printing on glass substrates (76 × 52 mm 2 , Paul Marienfeld GmbH & Co. KG) prepared with a sacrificial layer. First, the glass was cleaned by ultrasonication in 1:1 vol% 2-propanol:ethanol for 20 min (Branson 5510), after which the residual solvent was removed with N2 gas flushing. Next, a 10 wt% polyvinylpyrrolidone (PVP) in ethanol was cast on a glass substrate during spin coating (Karl Süss RC8). The spin coater was set to 3000 rpm, 300 rpm s -1 acceleration, and 45 s spin time. After spin coating, the treated substrates were post-treated at 100°C for 10 min.

Direct Ink Writing
A Hyrel EHR 3D printer equipped with a TAM-15 high-operating temperature syringe extrusion head was used for the controlled deposition of the thermoplastic PTU LCE. For this, a stainless-steel ink syringe and a 0.3 mm diameter micronozzle (Micro-Swiss MK8 M2549-03 3D) were used. The printing procedure is started by cutting the PTU LCEs into small chunks and drying them at 40 °C under a vacuum overnight. Subsequently, the syringe is loaded with the dried material and heated up to 200 °C for 20 min to ensure complete melting. Extrusion of the thermoplastic melt with the found settings induces molecular order parallel to the printing direction. The remaining material inside the syringe is emptied into a PTFE dish following the same print settings, mimicking the actuator formation while minimizing the exposure to PVP. For the recycling procedure, pristine printed actuators are cut into small pieces and reloaded into the syringe without further treatments to recycle the pristine printed material. Finally, the printing procedure is repeated as described previously to fabricate recycled actuators (1 st -4 th recycle). It should be noted that the crude material was heated for seven hours at 200 °C to yield the initial pristine actuators, and additional two hours of heating was required for each printing recycle, resulting in prolonged heat exposure of 15.5 hours after printing the complete recycling set. The actuation properties, thermal degradation, and alignment of the resulting prints were characterized using ImageJ image processing techniques, GPC, and X-ray scattering measurements, respectively. All print path files were obtained by manually writing the .g-code according to the desired printing conditions (speed, layer height, line spacing, and infill settings). With this technique, free-standing and patterned actuators were printed.

Patterned actuators
Patterned actuators are created by printing on 5 and 10 µm uniaxially stretched polyetherimide (PEI) foils with the reported settings. The PEI films are passive substrates that do not act as an alignment layer for oligomeric mesogens. 2 Two different pattern actuators are created with the produced .gcode files. The first is a 2 × 35 mm 2 (w × l) PTU LCE film deposited on top of a 5 × 35 mm 2 (w × l) PEI film. After deposition of this film, laser cutting was employed to extract a 5 × 25 mm 2 (w × l) film (40% coverage). 2 In addition, a 2 × 35 mm 2 (w × l) LCE film is deposited on a star-shaped PEI foil. This PEI foil is laser-cut according to a manually written vector path.

Laser Cutting
Precise cutting of the LCE patterned PEI foils was performed using a Laser Engraving Machine (CO2 Laser Plotter 50W 40 × 40, XM-3040), equipped with a 50 W max. CO2 laser beam. The cutting path used to slice the samples was defined using the RDWorksV8 V8.01.18 software with minimal (7.5%) and maximal (8.3%) laser intensities and cutting speed (300 mm s -1 ) set. After creating the desired cutting path, the file can be saved into a .rd (R documentation) file that can be loaded into the laser engraving machine.

Optical Profiling
The layer thickness of the multilayer films was measured with an optical profiling system (Sensofar S neox 3D Optical profiler, equipped with a 10× objective). With the profiling system, 3D profiles were recorded over the sample width of four consecutive samples yielding the average printed thicknesses.

Actuation performance
The thermal actuation was performed by placing the printed samples on a black aluminum sheet on a hotplate. The hotplate is heated from 30 to 110 °C with 10 °C temperature intervals. After reaching 110 °C, the samples were allowed to cool back to room temperature by removal from the hotplate. One full heating and cooling cycle was performed before the actuation measurements to eliminate any possible thermal history of all prints. Subsequently, three heating-cooling cycles of each printed actuator were averaged to obtain the reported actuation strain. Photographs of the samples are made with an Olympus OM-D E-M10 Mark III (with M.Zuiko 60 mm f/2.8 macro lens) in manual mode, and the samples were exposed to a LED light source for proper lighting. The actuation of all free-standing actuators is calculated on these photographs via ImageJ calculations. For this, the measurement of the reference sample sets the scale.  Figure S1. The synthesis of segmented PTU, using sequential addition reactions with reactive mesogens 1 (n=1, RM257) and 2 (n=2, RM82).  Intensity / a.u.

Supporting Tables and Figures
q / nm -1 Figure S8. 1D MAXS diffractogram of the crude material showing a peak in the small-angle region at q = 0.29 nm -1 arising from the interdomain spacing between the microphase-separated domains. Additionally, a peak is observed at q = 1.39 nm -1 due to the spacing between the soft segment chains.  Figure S9. Schematic of the programmed printing path designed to make six actuator stripes (35 × 2 mm 2 ).