4D Printed Light-Responsive Patterned Liquid Crystal Elastomer Actuators Using a Single Structural Color Ink

round of irradiation with on-and off-periods of 30 s using either a 455 or 365 nm LEDs. The luminescent devices used as a detector for the λ = 532 nm laser are doped with 0.5 wt.% fluorescent dye 7 or 8 , made in a procedure described in detail elsewhere. [46] Thermal actuation of the film was performed inside an oven over a temperature range of 25–130 ° C. Samples were suspended in the air using tweezers and filmed using an Olympus OM-D E-M10 Mk III with M. Zuiko ED 60 mm f /2.8 macro lens. Combined photographs showing actuators at rest and in actuated states were made by stacking the separate photographs in Adobe Photoshop CC using the “lighten” blending mode. The temperature during thermal actuation was simultaneously recorded with a Sensirion SHT3x thermocouple.

Envisioning robotic devices with functionality resulting from processing rather than exclusively material composition, multimodal responsive devices produced from a single azobenzene-functionalized cholesteric liquid crystal elastomer ink are demonstrated.The resulting device displays simultaneous structural color and actuation in response to both ultraviolet and blue light exposures.Through direct ink writing, a microextrusion-based additive manufacturing technique, the liquid crystalline ink is deposited into different effective mesophases-planar and slanted cholesteric, and a uniaxial pseudo-nematic state-by altering only two printing conditions: deposition speed and time before photopolymerization.Comparing the three different printed mesophases, marked differences in optical properties and actuation behaviors are found.Demonstrators based on bilayer architectures present concepts for interactive soft multiactuators deposited in a single printing step from a single ink.

Introduction
Soft robots comprised of flexible components can offer passive and resilient deformations to adapt to different shapes and environments, performing continuous and wide ranges of motion. [1]This makes them of interest in other situations that require adaptability and compliance, [2] or as alternatives to hard, metallic robots for applications where direct interaction with living creatures are required, such as in biomedical technology, where biocompatibility and damage limitation are of paramount importance during treatment. [3]Consequently, untethered soft robots are already being developed to realize minimally invasive surgeries and diagnoses. [4]

Synthesis of a Photo-Responsive Cholesteric Liquid Crystal Ink
The goal of this work is to simplify the production of photoresponsive, structurally colored, soft robotic actuators by developing a printing procedure capable of depositing multiple distinct nematic LC alignments through modulation of the azo-ChLC ink's supramolecular organization-uniaxial planar pseudo-nematic ("N"), planar cholesteric ("Ch"), and slanted cholesteric ("sCh")-during a single, uninterrupted print job, thus creating single actuators with multiple functionalities from one LC ink.The key is depositing the actuator without changing the temperatures of either the print head or substrate.We will discuss the two stages of development: ink synthesis and its deposition via DIW.
To achieve the viscoelastic ink capable of aligning into different mesophases, LC diacrylates and a dithiol (see Figure 1a for a depiction of all molecular structures) are concatenated into oligomers of varying lengths.For oligomerization, we use a widely-applied one-pot thiol-acrylate "Michael" addition reaction. [15]Forming the bulk of the ink is the common nematic diacrylate 1. Adding 8.7 wt.% of chiral dopant 2 allows generation of the Ch and sCh phases with its helical twisting power, HTP, in this mixture being HTP = 38.8µm −1 (see Figure S1, Supporting Information and ref. [22]) with visibly greenish reflection when used at the given concentration.To allow photo-induced actuation, 2 wt.% of the azobenzene-derived molecular photoswitch 3 was added.Dithiol 4 is chosen as the chain extension agent and organic base 5 is used in catalytic amounts to bring about the thiol-acrylate reaction (see Figure 1b for a stylized depiction of the procedure).The average number of mesogenic repeat units per oligomer ( LC x ) and the end-group functionality are controlled by the reactant stoichiometry. [15]A relatively low LC x is required to maintain control over the mesophase formation-too long oligomers impair the formation of the helicoidally-aligned cholesteric phases. [17,20,24]The molar ratio between the diacrylates 1+2+3 to dithiol 4 was set to 1.25 to 1, for a theoretical LC x = 5.An LC x = 4.3 is determined from 1 H-NMR data (see Figure S2, Supporting Information), corresponding to an M n ≈ 3440 g mol −1 with chain length dispersity Đ = 2.3 (leading to a calculated M w ≈ 7920 g mol −1 , determined using gel permeation chromatography, GPC), similar to values found in the literature. [23,24]he nematic-to-isotropic transition temperature, T NI , was determined by DSC to be T NI = 55 °C, as the peak of a broad transition signal (Figure S3, Supporting Information).

Design of a Multi-Mesophase Print Procedure
With the ink generated, the next step is developing efficient printing procedures capable of achieving different molecular alignments with the same ink, while maintaining DIW print parameters as similar as possible.We start optimization by determining conditions for achieving a planar cholesteric alignment with maximum reflectance and minimal scattering by printing on glass coated with a thin sacrificial layer of poly(vinyl alcohol) (PVA), which is subsequently dissolved by placing the samples in water, releasing the prints and forming freestanding films. [25]The DIW involves optimizing the temperature of the ink reservoir, T syringe , and the speed of the print head, v nozzle , which both influence aligning forces during extrusion, and finally the temperature of the printing bed, T bed , shown in previous work as vital to print quality. [13,16]fter screening several parameters, the printer-to-substrate nozzle gap was set to 50 µm (d LCE , Figure S4, Supporting Information).The effects of T syringe and T bed on the cholesteric reflectance are studied simultaneously: T syringe was varied between 40-90 °C and T bed set to 38, 44, or 56 °C (see Figure S5, Supporting Information), as outside of these ranges, the desired alignment characteristics were not found.Printing at v nozzle = 1 mm s −1 , which was found to be functional in previous work, [13] combined with T syringe = 70 °C and T bed = 44 °C appears to induce the greatest reflectance, R = 40% to 50% at λ max = 524 nm, with minimal scatter.The time between printing and photopolymerization, t wait , which allows thermodynamically driven formation of the helical twist of the cholesteric phases on the bed, was also found to be important: t wait ≈15 min on the printer bed is needed for proper formation and alignment of the cholesteric phase (Figures S6 and S7, Supporting Information).An additional consideration is that since azobenzene 3 isomerizes to the cis-form by absorbing 405 nm light used for photopolymerization (Figure S8, Supporting Information), a λ = 455 nm LED is simultaneously applied after the print during photopolymerization to drive the azobenzene to the predominantly trans population to avoid extraneous internal stresses. [26]Depositing the same ink at faster speeds (v nozzle = 10 mm s −1 ) using identical temperatures forms the sCh alignment (Figure 2a). [16]The resulting films displayed some variation in average thickness, roughly between 30 and 65 µm (the polymer substrates were 10 µm thick), with the samples being generally thinner at faster deposition rates.These thicknesses are sufficient to actuate such thin substrates. [24]aving established that the azo-ChLC oligomer ink can form the planar Ch and slanted sCh mesophases by varying only the printing deposition speed, we wished to determine if the cholesteric ink can be trapped in an otherwise unfavored uniaxial planar pseudo-nematic alignment.The time between printing and polymerization t wait is found to be the critical parameter in trapping the pseudo-N alignment, which is essentially a severely deformed cholesteric with the directors of the helical plains (nearly) parallel; it was successfully achieved by simultaneously printing at v nozzle = 10 mm s −1 and initiating photopolymerization immediately upon deposition of the ink (that is, an effective t wait ≈ 0).
Figure 2b displays polarized optical microscopy (POM) images of the Ch, sCh, and pseudo-N aligned prints.All alignments exhibit birefringence, which in the cholesteric samples appears equal irrespective of the crossed polarizer orientation.The sample printed with a high v nozzle and immediately polymerized (t wait ≈ 0), exhibits large differences in transmittance between 0 and 45°, indicating the pseudo-uniaxial molecular alignment.The apparent planar pseudo-nematic order indicated by the POM is further supported by wide-angle X-ray scattering (WAXS) data.Order parameters, S, have been calculated from the 2D X-ray diffractograms for all three alignments shown in Figure 2c. [27,28]While the LCs are themselves aligned in a nematic order on the molecular level, the effective bulk order parameters are S = 0 for Ch and sCh (Figure S9a,b, Supporting Information).For the strongly distorted pseudo-N aligned LCEs, uniaxial nematic order is also found in bulk at S = 0.31 (Figure S9c, Supporting Information), close to those found for similar DIW 3D-printed uniaxially aligned LCE materials. [29]A summary of the reflectance of the ink printed under differing conditions is seen in Figure 2d.
As a control experiment, we printed the pseudo-nematic alignment with bed temperature lowered to room temperature (RT).This lower bed temperature resulted in a slightly higher order parameter (S = 0.39, Figures S9d and S10, Supporting Information).While printing with T bed at RT-as often employed in reference DIW procedures [25,29,30] -offers somewhat greater pseudo-nematic order by easing the kinetic trapping of the N alignment, temperature adjustments dramatically increase production times if multi-mesophase prints are desired.Therefore, the slightly lower pseudo-nematic order obtained at T bed = 44 °C is well-compensated by an overall shorter device fabrication time.The inset shows the order parameter, S, calculated from this data using the "Kratky" method. [31]d) UV-vis reflectance spectra for Ch, sCh, and uniaxially planar pseudo-N alignments.The approximate absorbance band location for the azobenzene derivative is indicated: it accounts for the apparent negative reflectance values recorded.

Photo-Induced Mechanical Response of the Azo-Functionalized ChLCE
The three alignments, pseudo-N, Ch, and sCh, were also processed as bilayers by depositing the LCE directly on 10 µm thick PEI films. [24]By only varying v nozzle and t wait , any of the three alignments can be printed on this substrate.
We anticipated different actuation behaviors between the three alignments.Greater actuation strain is generally expected for LCEs with greater anisotropic order. [32]In LCE-PEI bilayer devices, during temperature increase, the difference in thermal expansion coefficients between the LCE and substrate leads to out-of-plane deformations (visualized in Figure 3a). [21,24,33]The Ch and pseudo-N LCEs have significantly different macroscopic orientations; the actuation of ChLCEs relies not on uniaxial but biaxial contraction [34] in the same plane as the PEI substrate.
We characterized the actuation modes of the differently aligned LCE prints (each 30 × 6 × 0.05 mm 3 , see Figure 3b,c).All photoactuation experiments were performed at RT using either λ max = 365 nm or λ max = 455 nm LEDs at 15 cm distance.In all LCE-PEI bilayer photoirradiation experiments, the LCE side of the bilayer sample is irradiated.7] The photoactuation modes (see Figure 3d for the experimental setup) of all alignments irradiated with 365 are shown in Figure 3e and Video S1, Supporting Information (similar actuation is seen when the films were exposed to either 455 nm or simultaneously to 455 and 365 nm, see Video S2, Supporting Information).The bilayer Ch and pseudo-N prints contract on the exposed side and bend toward the light, independent of the LED wavelength, with the pseudo-N demonstrating greater and faster actuation, likely a result of increased molecular alignment.Irradiation of the latter with the 365 nm LED from slightly below can cause self-shadowing, resulting in dramatic oscillatory motion (Video S3, Supporting Information). [40]n contrast, the slanted cholesteric samples bend away from the excitation light.This "reverse" deflection is explained by considering the thermomechanical forces generated, as illustrated in Figure 3a.At θ slant = 0°, the exposed surface contracts in-plane, [41] and the bilayer device bends toward the light source; at a hypothetical θ slant = 90°, expansion along the helix director is expected, and the bilayer device should bend away from the light source. [42]etween these two extremes, we anticipate a transition.We hypothesize that since the nematic planes are at an angle θ slant with respect to the substrate surface, the regular contraction is accompanied by an expansion parallel to the PEI substrate.At an unspecified θ slant , the material theoretically will reach a state of equilibrium between contraction and expansion parallel to the PEI, resulting in no bending.As the θ slant increases above a critical angle, the helical axis is aligned more parallel to the PEI substrate.In such samples, as expansion parallel to the PEI dominates, the bilayer film bends away from the light, so we postulate the bending sCh LCE must have a θ slant greater than the "equilibrium state".Visual inspections of the green reflectance of an sCh film tilted at different angles demonstrate it reflects most intensely between 45 and 60°, which after taking light refraction into account, would suggest 27° < θ slant < 34°. [22]The exact tilt for this system has not been elucidated through other means.
From the looped print pattern, it may be expected that "2D-to-3D" out-of-plane deformations happen as shown previously for circular alignment patterns. [25,38,39]However, as the print paths here are printed in a clockwise fashion, the aspect ratio l/w up to 5 would naturally lead to a contraction along the "l-axis" that is much more pronounced macroscopically than the deformation arising from the small sections printed along the w-direction.Another complicating factor is the bonded PEI substrate, which allows for the formation of curvature only along a single plane (i.e., bending, as opposed to cone or saddle formation).
We find a strong correlation between incident light intensity and tip displacement for both 365 nm and 455 nm LED exposures (Figures S11-S14, Supporting Information) for a ChLCE-PEI bilayer.After disengaging the light source, the sample deformation quickly reverts to the initial state regardless of the initial illumination intensity, hinting that the photomechanical stresses induced by trans-to-cis isomerization of the azobenzene moiety are not themselves sufficient to maintain the bent state after removal of the light source.Concurrently, significant sample heating is found, indicating that the mechanism of actuation is mainly reliant on a photothermal effect, as seen in other nematic "4D printed" azo-LCEs. [20]High intensities of excitation light result in high temperatures, so a thermogravimetric analysis (TGA) was done on the crosslinked film to gauge thermal stability (Figure S15, Supporting Information).The material does not lose mass until at least 300 °C, indicating the temperatures reached during actuation, up to 130 °C, do not directly damage the material.

MultiModal Bilayer Actuators Azo-Functionalized ChLCE
The goal now was to print actuators with multiple elements in a single printing action, to demonstrate multimodal actuations are possible.The first multifunctional actuator discussed is an LCE-PEI bilayer containing all three alignments as individual 10 × 6 × 0.05 mm 3 segments, each printed in a single print run using the same ChLC oligomer ink (Figure 4a).As can be seen in Figure 4b, the difference in actuation characteristics of the three distinct mesophase domains in the device allow intricate bending deformations upon illumination (see Video S4, Supporting Information for a real-time demonstration).We make use of these multiple bending deformations and photonic nature of the various segments in the example multifunctional device shown in Figure 4c.Behind the suspended film and within the path of the excitation laser, we place a polymer plate doped with a fluorescent dye (7, see Experimental) that absorbs the incident laser light and emits bright red fluorescence (λ em = 630 nm), effectively acting as "scintillator" for the laser beam.Initially, the 532 nm laser light incident on the Ch section is strongly reflected so almost no light reaches the fluorescent plate, which remains dark (Figure 4c, top).However, when the sCh region (only) is actuated with λ = 365 nm LED light (irradiance 400 mW cm −2 , Figure 4c, middle) the film starts to bend and rotate, changing the incidence angle of the laser light to the Ch segment, now allowing some laser light to pass through and strike the fluorescent plate, which appears as a faint red dot in the photograph.When the pseudo-N region at the top of the suspended film is then actuated (Figure 4c, bottom), the resulting displacement removes the Ch segment entirely from the laser beam, which now falls directly on the fluorescent plate, resulting in a bright red illumination spot.Thus, the effective positioning of the actuation mode can be visualized and monitored via the fluorescent detector plate.
A second functional example relies on selective reflectance rather than selective transmission.The photonic actuator made in a single printing job displayed in Figure 4e is divided into planar pseudo-N (20 × 6 × 0.05 mm 2 ) and Ch segments (10 × 6 × 0.05 mm 2 ).Suspended in air, the planar pseudo-N segment is actuated with 365 nm UV light while a laser shines λ = 532 nm light on the Ch segment from the same side.The laser light reflects from the Ch segment.By adjusting the intensity of the UV light on the pseudo-N segment, the film is actuated, the degree of bending correlating to the illumination intensity.The bending changes the orientation of the Ch segment, which now reflects the incident light toward the luminescent detector.As a potential application of such a construct, one can imagine replacing the fluorescent plate with a digital photodetector and controller for adjusting the UV light source intensity.Illumination of the pseudo-N segment creates a feedback loop where the film alternately unbends and bends, turning "off" and "on" the UV excitation light, respectively, triggering a continuous back-and-forth out-of-equilibrium motion of a photo-responsive polymer film that does not rely on self-shadowing effects.

Conclusions
Multicomponent actuators are produced in one DIW 3D print step from a single ink by varying only the deposition speed and delay time before photopolymerization.Combining a properly developed ink and deposition method greatly simplifies the generation of multifunctional actuators, allowing fabrication in significantly shorter timeframes.By depositing all the segments at the same processing temperature, one of the greatest time elements, that is, the time taken to adjust and steady the absolute and relative temperatures of the print head and the printing bed, are essentially eliminated.The potential application of these complex actuating photonic devices is exhibited in two demonstrators which interact with external light sources to turn "on" and "off" transmissive or reflective properties of the photonic segments by physical manipulation of the film orientation by the actuator segments.We anticipate these demonstrators will inspire researchers to develop increasingly intricate "communicating" actuators capable of interacting with and responding to their environment that may be easily generated in a single print pass with a single ink.
Azo-ChLCE Oligomer Ink Synthesis: Solid components 1, 2 (8.7 wt.%) and 3 (2 wt.%) were added in a 5:4 weight ratio with 4 and dissolved in 2 mL CH 2 Cl 2 per 1 mg solid to a 50 mL round-bottom flask and heated to 39 °C.After 30 min of magnetic stirring, the CH 2 Cl 2 dissolved oligomerization catalyst 5 was added to the monomer mixture via pipette and the flask was closed using loose-fitting aluminum foil.After 24 h the reaction was completed by removing the stirring bar and turning off the magnetic plate and heater.The oligomers were dissolved in CH 2 Cl 2 and washed twice in a 100 mL separation funnel with 1 M HCl to remove 5.The organic phase was dried with MgSO 4 and separated from the drying salt under vacuum with a Büchner flask.The CH 2 Cl 2 was evaporated in a vacuum oven at 85 °C (above the T NI of the oligomers) for an hour.For analytical purposes, 0.2 g of solid content was extracted from the oligomer mixture.Then, the oligomer mixture was dissolved in CH 2 Cl 2 and photoinitiator 6 was added.The resulting mixture was stirred for another 30 min.The CH 2 Cl 2 was again evaporated inside a vacuum oven at 85 °C for an hour.Afterward, the ink is transferred to a DIW cartridge to be used for printing.
Ink Characterization: To assess the success of the oligomerization reaction, 1 H-NMR spectra were measured from the oligomeric analytical sample taken.The sample was dissolved in chloroform-d (purchased from Sigma-Aldrich Inc., 99.8 atom% D, 0.03% v/v tetramethylsilane) and recorded with a Bruker Avance III HD 400 MHz spectrometer: see Supporting Information for details.The average number of mesogenic repeat units per oligomer was calculated using (8.0 8.5) (7.9 8.0) (7.8 7.9) 4 after normalizing the acrylate peaks (see Figure S2, Supporting Information): the peaks within the chemical shifts (δ, ppm) 8.0-8.5, 7.9-8.0,and 7.8-7.9ppm represent four specific protons per mesogenic core of the oligomerized monomers 1, 2, and 3, respectively. [13,24]To assess the dispersity Đ of the ink, GPC plots were recorded with a Shimadzu LC-2030.3D with a 254 nm photodiode array and refractive index detectors using tetrahydrofuran as eluent (purchased from Biosolve B.V., stabilized with butylated hydroxytoluene).The T NI and T g of the oligomeric mixtures were quantified using differential scanning calorimetry (DSC) performed with a TA Instruments Q2000 within a temperature range of 50-120°C with a rate of 10 °C min −1 for both the heating and cooling ramps (see the supporting information). [43]irect Ink Writing Procedure: Three substrates were used in this work: glass, PVA-coated glass and 10 ± 2 µm thick polyetherimide foil (Ultem UTF 120, purchased from SABIC) taped on glass, called "PEI" throughout this work.The borosilicate glass used for the glass and PEI substrates was rinsed with acetone and blown dry using an N 2 air gun.The final step in PEI substrate preparation was flattening and taping the plastic onto the glass substrate.The borosilicate glass for the PVA coated glass substrate was prepared by ultrasonicating the glass for 20 min in a 1:1 v/v ethanol:2-propanol solution using the Branson 2510 Ultrasonic Cleaner, followed by 20 min in a UV/O 3 (UV Products PR-100).PVA in a 4% w/w solution with deionized H 2 O was applied to the glass at 1500 rpm for 30 s using a Karl Suss RC6 spin coater.Prints were made using a Hyrel Engine HR equipped with TAM-15 high-operating temperature reservoirs.Micronozzles with 0.335 mm diameter (Fisnar QuantX Micron-S Red) were attached to the ink-containing cartridge via a Luer-Lock adapter.Temperatures of the print head, T syringe , and the print bed, T bed , were measured using a Kane-May KM330 digital thermometer.The G-code file was generated using a homebrew Python-based tool utilizing the open-source app framework "Streamlit", which writes concentric print paths and processes the parameter values chosen by the user to print simple lines or more complex shapes such as rectangles, circles, cubes, cylinders, pyramids, or cones.The complete code is available free of charge online. [44]The printed structures were initially photopolymerized at T bed using LEDs with wavelengths of both 405 nm (distance 40-50 cm, 0.15 mW cm −2 in UV-A range) and 455 nm (distance 10-15 cm, 0 mW cm −2 in UV-A range) and after completion of the print, removed to an N 2 atmosphere and exposed for 30 min per side under using a highintensity UV light source (Excelitas EXFO Omnicure S2000) with a > 400 nm cut-on wavelength filter.UV intensity was measured to be 20 mW cm −2 (range 330-455 nm) and 1-2 W cm −2 (range 315-400 nm); intensity data collected with an Opsytec Dr. Gröbel RM-12 Radiometer (0-199 mW cm −2 range, 0.01 mW cm −2 resolution) using both UV-A+ ("RM-12" detector, range 330-455 nm) and UV-A (range 315-400 nm) photosensors.
Printed Azo-ChLCE Characterization: Gel fraction f gel of polymerized films were measured by immersing in THF and soaking for 48 h, followed by subsequent drying in a vacuum oven at 85 °C for 1 h.Subsequently, the gel fraction was calculated as f gel = m final /m initial .The prints were visually analyzed and photographed, and the reflective properties were inspected using POM in reflectance mode with crossed polarizers with a Leica DM2700M microscope equipped with a Leica DFC 420C camera in transmittance mode to investigate the nematic order of the LCEs.X-ray scattering measurements were performed using a SAXSLAB Ganesha 300XL X-ray diffractometer with a Genix-Cu X-ray source (Cu Kα, λ = 0.154 nm) and a Pilatus 300K detector (resolution 487 × 619, pixel pitch 172 µm).Scattering vectors were calibrated against a silver docosanoate standard.The collected data were reduced and analyzed using a custom Python script with the "PyFAI" software package.Estimation of the molecular order parameter, S, was done using the "Davidson" and "Kratky" methods. [31,45]The surface profiles of the prints were evaluated using a Sensofar S neox 3D optical profiler.Samples were placed onto the measuring table to extract the film thickness z and surface profiles of the prints (see Figure S16, Supporting Information for an example).Reflectance, absorbance, and transmittance spectra were obtained using a PerkinElmer Lambda 750 spectrophotometer with 150 mm integrating sphere detector.To measure the influence of the photoisomerization of azobenzene 3 on the transmissive and reflective properties of a film, photoisomerization was done using λ = 365 nm and 455 nm LEDs in a dark environment.The respective light intensities at 10 cm distance were 13 mW cm −2 (365 nm LED) and 20 mW cm −2 (455 nm LED).Samples were placed at 10 cm distance from the LEDs and irradiated for 3 min with 365 nm light, followed by immediate UV-vis measurements.Back-isomerization was performed with 455 nm light irradiation for 5 min.Finally, transmittance at oblique angles was measured using the OMT Solutions ARTA goniometer sample stage detector instead of the integrating sphere detector.The sample was tilted along its width over a range of −75 -75° in increments of 15°.TGA measurements were performed using a TA Instruments Q500 instrument on a 4 ± 0.5 mg print sample over a temperature range from 25 to 600 °C and a heating rate of 10 °C min −1 .The point of thermal decomposition was characterized using the "onset function" of TA Instruments TRIOS software.
Printed Device Actuation Characterization: The LCE was removed from glass substrates using water to dissolve sacrificial PVA layers or via cutting the printed area from the unprinted area of the plastic substrate, respectively, and were suspended from tweezers in the air with black background at 20 °C.365 and 455 nm LEDs were stationed at 15 cm distance.Photothermal and photomechanical actuation was characterized using the 365 nm LED and photothermal actuation with the 455 nm LED.The actuation was filmed using an Olympus OM-D E-M10 Mk III with M. Zuiko ED 60 mm f/2.8 macro lens.This camera was used in conjunction with a Xenics high-speed IR camera to relate tip displacement to film temperature of LCEs during photoirradiation.The input intensity was increased from ≈25-425 mW cm −2 each round of irradiation with on-and off-periods of 30 s using either a 455 or 365 nm LEDs.The luminescent devices used as a detector for the λ = 532 nm laser are doped with 0.5 wt.% fluorescent dye 7 or 8, made in a procedure described in detail elsewhere. [46]Thermal actuation of the film was performed inside an oven over a temperature range of 25-130 °C.Samples were suspended in the air using tweezers and filmed using an Olympus OM-D E-M10 Mk III with M. Zuiko ED 60 mm f/2.8 macro lens.Combined photographs showing actuators at rest and in actuated states were made by stacking the separate photographs in Adobe Photoshop CC using the "lighten" blending mode.The temperature during thermal actuation was simultaneously recorded with a Sensirion SHT3x thermocouple.

Figure 1 .
Figure1.a) Reactants for synthesis of the cholesteric oligomer ink; the reactive mesogen 1, chiral dopant 2, and azobenzene derivative 3; the dithiol chain extender 4; the base catalyst 5; and the photoinitiator 6, allowing free radical polymerization crosslinking after printing.b) Idealized illustration of the "molecular situation" during the different steps of handling the liquid crystalline ink, from oligomer synthesis to crosslinking after direct ink writing.

Figure 2 .
Figure 2. a) Idealized illustrations of the various (chiral) nematic alignments achieved.b) Polarized optical microscopy images of printed samples aligned with printing direction parallel (0°) or at an angle (45°) with respect to the polarizers.c) Wide-angle X-ray scattering data for free-standing films at different alignments.The inset shows the order parameter, S, calculated from this data using the "Kratky" method.[31]d) UV-vis reflectance spectra for Ch, sCh, and uniaxially planar pseudo-N alignments.The approximate absorbance band location for the azobenzene derivative is indicated: it accounts for the apparent negative reflectance values recorded.

Figure 3 .
Figure 3. a) Expected in-plane deformations of the ChLC layer for a variety of theoretical cases; that is, planar cholesteric, slanted cholesteric, or a homeotropic cholesteric alignments.b,c) The geometry of the print path design for analysis samples, with the physical dimensions, as well as the print path direction: in (b), the "top view" is illustrated, in (c), the cross-section.d) Experimental setup for the irradiation experiments.A collimated LED light bundle is aimed at the ChLCE-PEI bilayer device from 15 cm.e) Final actuation states for ChLCE-PEI bilayers, for different irradiation regimes.In all cases, the ChLCE side of the actuator is irradiated.The white curves are added to aid interpretation of the photographs.

Figure 4 .
Figure 4. a) Photograph of the multi-mesophase film printed on PEI foil.Indicated are the locations of the uniaxial pseudo-nematic, slanted cholesteric, and planar cholesteric domains.b) Combined photographs highlighting complex actuation behaviors using the different mesophase domains.The film is irradiated with both λ = 365 nm and λ = 455 nm lights simultaneously.c) Illustrative comparison showing how the reflective element on the actuator can be used to intercept, wholly or partially, a λ = 532 nm laser beam, or leave it to pass untouched.The left column photographs show a luminescent device using fluorescent dye 7 acting as an optical sensor for the incident laser beam.The photographs in the right column serve purely illustrative purposes.d) Photograph of an alternative device design, featuring a large uniaxially planar pseudo-nematic section for actuation or light steering, and a planar cholesteric section for light reflection.e) Demonstration of irradiance-dependent light steering.A λ = 532 nm laser is directed towards the reader at the planar cholesteric section and obliquely reflected to a luminescent device at a distance.An λ = 365 nm LED (intensity indicated in each image) is used to actuate the ChLCE films and redirect the incident laser light.Markers are added to indicate the path of the laser (dotted line), LED light direction (solid green line), and position of the reflected laser spot on the luminescent devices (dotted white circle: top fluorescent plate uses dye 7, bottom uses dye 8).NB: the scale bar in (a) and (d) corresponds to w = 6 mm.