Synchrotron Microbeam Diffraction Studies on the Alignment within 3D-Printed Smectic-A Liquid Crystal Elastomer Filaments during Extrusion

: 3D printing of novel and smart materials has received considerable attention due to its applications within biological and medical ﬁelds, mostly as they can be used to print complex architectures and particular designs. However, the internal structure during 3D printing can be problematic to resolve. We present here how time-resolved synchrotron microbeam Small-Angle X-ray Diffraction ( µ -SAXD) allows us to elucidate the local orientational structure of a liquid crystal elastomer-based printed scaffold. Most reported 3D-printed liquid crystal elastomers are mainly nematic; here, we present a Smectic-A 3D-printed liquid crystal elastomer that has previously been reported to promote cell proliferation and alignment. The data obtained on the 3D-printed ﬁlaments will provide insights into the internal structure of the liquid crystal elastomer for the future fabrication of liquid crystal elastomers as responsive and anisotropic 3D cell scaffolds.


Introduction
Synchrotron X-ray microbeam diffraction has been widely used to correlate optical microscopy images with local X-ray structure. One of the earliest studies was done on beamline 4A at the Photon Factory to characterize the local layer structure across a broad wall in a surface-stabilized ferroelectric liquid crystal (FLC) [1]. Other X-ray microbeam liquid crystal (LC) studies include the time-resolved study of an FLC [2], elucidation of the structure of LC phases derived from bent-core molecules [3], a combined X-ray microbeam, polarized optical microscopy, a scanning electron microscopy study of the internal structure of periodic smectic-A (Sm-A) zig-zag defects from mesoscopic scale to molecular level [4], and the determination of orientational periodicities in the LC smectic-C* (Sm-C*) subphases by using an X-ray energy at the resonant edge of selenium-containing compounds [5]. One of the authors (RP) performed his first X-ray microbeam experiment with Professor Clark's group to correlate polarized optical microscopy images with local X-ray microbeam diffraction from short DNA oligomers [6]. This was a highly memorable project that was later highlighted in an engaging book [7]. In the current work, X-ray provides an entirely new platform to study the 3D printing of Sm-A LCE elastomers and evaluates the effect of extrusion using synchrotron x-ray microbeam diffraction.
Amber syringe barrel, white polyethylene piston, end cap and blue dispense tip with an internal diameter of 0.41 mm (Nordson PN 7018272) were purchased from Nordson.
The print was obtained with our in-house 405 nm photosensitive LCE-based ink containing Irgacure 819 photoinitiator (Bis (2,4,6-trimethylbenzoyl)-phenylphosphine oxide). A curing of 75 seconds was applied in situ followed by 15 minutes ex situ post-curing. The UV system consisted of an Omnicure S1500 UV lamp (200 W Mercury lamp with a 320-500 nm filter) at 75% intensity and used a light guide placed at 13 cm from the sample in situ and 20 cm during ex situ post-curing. The print was made onto a 25 µm thick polyimide substrate (CAPLINQ PN PIT2A).
Scanning electron microscopy (SEM) was performed on samples which were goldcoated (700 Å) using a sputter coater (Hummer VI-A, Anatech Ltd., Springfield, VA, USA) at 10 mA DC for 3 min. Images were obtained using Quanta 450 FEG SEM under high vacuum. 1 H and 13 C NMR of copolymers and elastomers were recorded in CDCl 3 at room temperature on a Bruker DMX 400 MHz instrument and referenced internally to residual peaks at 7.26 ( 1 H).

Synchrotron µ-SAXD
Measurements were performed in transmission mode of the Soft Matter Interfaces beamline (12-ID SMI) at the National Synchrotron Light Source II (NSLS-II) [49]. Scattering patterns were measured at 16.1 keV with a beam size at the sample position of 2.5 µm × 25 µm (V × H) and recorded on a Pilatus 300 K−W detector, consisting of 0.172 mm square pixels in a 1475 × 195 array, mounted at a fixed distance of 0.275 m from the sample position. The 2D raw SAXS images were acquired in-air and converted into q-space, followed by a fixed background subtraction. Local variations in thickness, liquid crystal concentration, and morphology may have impacted the baseline after subtraction of a fixed background. Reduced images were visualized in Xi-CAM software and radially and azimuthally integrated using a custom Python code [49].

Results and Discussion
In situ µ-SAXD and 3D printing experiments were performed on a customized 3D printing platform mounted in-line with the X-ray beam in the sample chamber of the Soft Matter Interfaces (SMI) beamline at NSLS-II. The printer design was detailed by Wiegart et al. [15]. The entire printer was placed on the sample stage equipped with full x, y, z translation to position the printing process in-line with the X-ray beam. An on-axis and wide-view camera were used to assist in the alignment. An extrusion barrel filled with LCE-based ink was positioned on the printhead and horizontally/vertically moved to print the ink across the substrate on the print bed (see Figure 1). LCE filaments were extruded through a 22 GA straight-tip metal nozzle with an inner diameter of 410 µm during translation of the barrel at 1.5 mm/s along the print bed onto a polyimide substrate. The distance between the nozzle and the substrate was kept constant at 300 µm. Constant extrusion pressure of 25 psi was controlled by a Nordson Ultimus V dispenser. during translation of the barrel at 1.5 mm/s along the print bed onto a polyimid The distance between the nozzle and the substrate was kept constant at 300 μm extrusion pressure of 25 psi was controlled by a Nordson Ultimus V dispense The processing parameters were optimized for the ink viscosity to achi printability and reproducible filament structure, confirmed by Scannin Microscopy (SEM) (Figure 2) performed on the printed filaments. The SE revealed that the filament experienced shape deformation at the substrat forming a flat bottom after deposition due to relaxation, viscoelastic effects a of the substrate. The flattening of the filament resulted in a reduction in the filam in the solidified structure to about 240 μm compared to the 300 μm height exp the settings of the nozzle height above the print bed. Overall, SEM imag consistent shape and a homogeneous, void-free texture across the filament sectional areas. The X-ray beam was oriented perpendicular to the printing direction. A system was focused onto the printing platform with an adjustable mounting ar distance of ~13 cm from the intersection of the X-ray beam and the print bed The illumination was oriented in the same direction as the horizontal printin experiment, the in situ printing and SAXS measurement protocol was as follow began 5 mm away from the intersection of the X-ray beam and was horizontally towards the beam (Figure 3). The UV curing system was turned on as soon as t commenced to initiate crosslinking and was turned off after 75 s curing ti printed filament intersected the X-ray beam, a positional trigger was used to si area detector to begin acquiring scattering patterns. The point at which th crossed the beam was designated as time = 0. The X-ray beam was located at The processing parameters were optimized for the ink viscosity to achieve smooth printability and reproducible filament structure, confirmed by Scanning Electron Microscopy (SEM) ( Figure 2) performed on the printed filaments. The SEM images revealed that the filament experienced shape deformation at the substrate interface, forming a flat bottom after deposition due to relaxation, viscoelastic effects and wetting of the substrate. The flattening of the filament resulted in a reduction in the filament height in the solidified structure to about 240 µm compared to the 300 µm height expected from the settings of the nozzle height above the print bed. Overall, SEM images show a consistent shape and a homogeneous, void-free texture across the filament and cross-sectional areas.
Crystals 2021, 11, x FOR PEER REVIEW 4 of 10 during translation of the barrel at 1.5 mm/s along the print bed onto a polyimide substrate. The distance between the nozzle and the substrate was kept constant at 300 μm. Constant extrusion pressure of 25 psi was controlled by a Nordson Ultimus V dispenser. The processing parameters were optimized for the ink viscosity to achieve smooth printability and reproducible filament structure, confirmed by Scanning Electron Microscopy (SEM) ( Figure 2) performed on the printed filaments. The SEM images revealed that the filament experienced shape deformation at the substrate interface, forming a flat bottom after deposition due to relaxation, viscoelastic effects and wetting of the substrate. The flattening of the filament resulted in a reduction in the filament height in the solidified structure to about 240 μm compared to the 300 μm height expected from the settings of the nozzle height above the print bed. Overall, SEM images show a consistent shape and a homogeneous, void-free texture across the filament and crosssectional areas. The X-ray beam was oriented perpendicular to the printing direction. A UV curing system was focused onto the printing platform with an adjustable mounting arm at a fixed distance of ~13 cm from the intersection of the X-ray beam and the print bed ( Figure 3). The illumination was oriented in the same direction as the horizontal printing. As a first experiment, the in situ printing and SAXS measurement protocol was as follows: printing began 5 mm away from the intersection of the X-ray beam and was horizontally translated towards the beam (Figure 3). The UV curing system was turned on as soon as the printing commenced to initiate crosslinking and was turned off after 75 s curing time. As the printed filament intersected the X-ray beam, a positional trigger was used to signal the 2D area detector to begin acquiring scattering patterns. The point at which the filament crossed the beam was designated as time = 0. The X-ray beam was located at a height of 150 μm above the print bed to characterize the structure formation in the approximate center of the filament. The microbeam sizing 2.5 × 25 μm (V × H) facilitated spatial scanning for a real-time structural analysis of the filament development. The X-ray beam was oriented perpendicular to the printing direction. A UV curing system was focused onto the printing platform with an adjustable mounting arm at a fixed distance of~13 cm from the intersection of the X-ray beam and the print bed ( Figure 3). The illumination was oriented in the same direction as the horizontal printing. As a first experiment, the in situ printing and SAXS measurement protocol was as follows: printing began 5 mm away from the intersection of the X-ray beam and was horizontally translated towards the beam (Figure 3). The UV curing system was turned on as soon as the printing commenced to initiate crosslinking and was turned off after 75 s curing time. As the printed filament intersected the X-ray beam, a positional trigger was used to signal the 2D area detector to begin acquiring scattering patterns. The point at which the filament crossed the beam was designated as time = 0. The X-ray beam was located at a height of 150 µm above the print bed to characterize the structure formation in the approximate center of the filament. The microbeam sizing 2.5 × 25 µm (V × H) facilitated spatial scanning for a real-time structural analysis of the filament development. Thirty SAXS measurements were taken continuously at an exposure time resolve the structure as a function of distance from the advancing printing fr structure could also be evaluated as a function of time after deposition with re time = 0. For each position, the radial and azimuthal μ-SAXD profiles were proces radial μ-SAXD profile exhibited two sharp scattering peaks at q1 = 0.16 Å −1 and Å −1 characteristic of a Sm-A organization with interdigitated cholesterol described in our previous work (see Figure 1 and Supplementary Information, S S1) [46][47][48]50]. μ-SAXD radial and azimuthal maps were extracted from the amp the scattering peak and the amplitude of azimuthal integration, respectively, aro 0.32 Å −1 (SI, Figure S1 and Figure 4, respectively). The Sm-A structure was fou persistent over the filament length. Figure 4 reveals the alignment of the Sm-A with respect to the horizontal printing direction, which was designated as a 0 azimuthal angle on the scattering detector. The orientation of the Sm-A diffractio (layer normal) appeared to be centered around 90° through the entirety of the process, indicating the long axis of the interdigitated moieties was aligned perpe to the printing direction. Therefore, the polymer chains would align along t direction to accommodate the LC side chain pendant configuration. The dist around the preferential angle in the azimuthal map was centered on the anisotropy associated with the extrusion process [26]. In comparison, a ring patt obtained in samples that were processed by film casting or particle leachin suggesting the preferential alignment was induced by 3D printing [46,50,51]. Ear filament development (time < 2.22 s), the orientation of the LCE layering w uniform, indicated by several individual streaks in intensity centered around scattering pattern was reminiscent of powder-like diffraction due to discrete g aligned Sm-A that formed in the filament and pointed to a distribution of d directions (see SI, Video S1). The flow was likely chaotic at these early times as was deposited from the high shear and elongational stress present in the nozz deposition and in the presence of the UV, the polymer chains began to crosslink, to an internal rearrangement, and a final structure was set in. After about 4 s (or a of 6 mm), the azimuthal intensity became more uniformly centered on 90°. Here Thirty SAXS measurements were taken continuously at an exposure time of 1 s to resolve the structure as a function of distance from the advancing printing front. The structure could also be evaluated as a function of time after deposition with respect to time = 0. For each position, the radial and azimuthal µ-SAXD profiles were processed. The radial µ-SAXD profile exhibited two sharp scattering peaks at q 1 = 0.16 Å −1 and q 2 = 0.32 Å −1 characteristic of a Sm-A organization with interdigitated cholesterol moieties described in our previous work (see Figure 1 and Supplementary Information, SI Figure S1) [46][47][48]50]. µ-SAXD radial and azimuthal maps were extracted from the amplitude of the scattering peak and the amplitude of azimuthal integration, respectively, around q = 0.32 Å −1 (SI, Figure S1 and Figure 4, respectively). The Sm-A structure was found to be persistent over the filament length. Figure 4 reveals the alignment of the Sm-A layering with respect to the horizontal printing direction, which was designated as a 0 or 180 • azimuthal angle on the scattering detector. The orientation of the Sm-A diffraction peaks (layer normal) appeared to be centered around 90 • through the entirety of the printing process, indicating the long axis of the interdigitated moieties was aligned perpendicular to the printing direction. Therefore, the polymer chains would align along the flow direction to accommodate the LC side chain pendant configuration. The distribution around the preferential angle in the azimuthal map was centered on the intrinsic anisotropy associated with the extrusion process [26]. In comparison, a ring pattern was obtained in samples that were processed by film casting or particle leaching foam, suggesting the preferential alignment was induced by 3D printing [46,50,51]. Early in the filament development (time < 2.22 s), the orientation of the LCE layering was non-uniform, indicated by several individual streaks in intensity centered around 90 • . The scattering pattern was reminiscent of powder-like diffraction due to discrete grains of aligned Sm-A that formed in the filament and pointed to a distribution of different directions (see SI, Video S1). The flow was likely chaotic at these early times as the LCE was deposited from the high shear and elongational stress present in the nozzle. After deposition and in the presence of the UV, the polymer chains began to crosslink, leading to an internal rearrangement, and a final structure was set in. After about 4 s (or a distance of 6 mm), the azimuthal intensity became more uniformly centered on 90 • . Here, the LC layering orientation aligned to the preferred direction induced by the print direction. However, a few streaks were still identified, possibly due to the isolated grains that solidified in a partially offset direction during the progression of the UV curing (see Figure S2). 3D printing is a complex non-equilibrium process: there can be high elongation and shear stress during extrusion from the nozzle, elongation in the 90 • bend when exiting the nozzle, flow transverse to the print direction before LCE is crosslinked due to substrate wetting and relaxation, additional elongation stress resulting from printhead translation, and inhomogeneous UV curing across the filament. This complexity makes it difficult to predict how the SmA LCE will orient. In situ measurements of the orientational ordering are, therefore, required to optimize the processing parameters (pressure, speed, nozzle diameter, height above print bed, ink viscosity, cross-linker concentration, UV light intensity) to optimize the alignment. Our measurements demonstrate that we have identified a set of processing parameters that, after an initial disordered stage, provide a Sm-A LCE filament with the Sm-A layers aligned approximately parallel to the print direction within ±5 degrees.
Crystals 2021, 11, x FOR PEER REVIEW nozzle, flow transverse to the print direction before LCE is crosslinked due to s wetting and relaxation, additional elongation stress resulting from printhead tra and inhomogeneous UV curing across the filament. This complexity makes it di predict how the SmA LCE will orient. In situ measurements of the orientational are, therefore, required to optimize the processing parameters (pressure, speed diameter, height above print bed, ink viscosity, cross-linker concentration, U intensity) to optimize the alignment. Our measurements demonstrate that identified a set of processing parameters that, after an initial disordered stage, p Sm-A LCE filament with the Sm-A layers aligned approximately parallel to t direction within ± 5 degrees. To expand the study of the orientation of the mesogenic moieties within structure, a second experiment consisting of a height scan was performed in the direction once the filament was completely printed and fully cured. The mi precision allowed for resolution of the SmA layering alignment in the solid structure. The X-ray beam was positioned at 5 mm corresponding to the midd filament development. The base of the filament was calibrated with respect to t beam (h = 0 μm). The printer was translated at discrete steps of 9 μm in the direction where at each point a SAXS image was taken, effectively scanning the s a function of height. Similar to the previous experiment, μ-SAXD radial and a maps were obtained, and the results are summarized in SI, Figure S3 and F respectively, by the change in intensity in the azimuthal integration at q = (integrated from q = 0.29 Å −1 to q = 0.35 Å −1 ) along the scanning height. Here, t orientation evolved spatially throughout the filament (see Video S2 and Figure influenced by the gradient in stress from the printing process as well as the penetration through the thickness of the filament. At the substrate interface (h ~ 0 azimuthal angle intensity distribution was found to be narrow, suggestin alignment of the LCE layering. This region underwent a higher shear stres printing compared to the bulk of the filament. As a consequence, the distrib orientations increased, progressing in the bulk of the filament. The distributio quantitatively determined from 1D line cuts of the azimuthal intensity at variou To expand the study of the orientation of the mesogenic moieties within the 3D structure, a second experiment consisting of a height scan was performed in the vertical direction once the filament was completely printed and fully cured. The microbeam precision allowed for resolution of the SmA layering alignment in the solidified 3D structure. The X-ray beam was positioned at 5 mm corresponding to the middle of the filament development. The base of the filament was calibrated with respect to the X-ray beam (h = 0 µm). The printer was translated at discrete steps of 9 µm in the vertical direction where at each point a SAXS image was taken, effectively scanning the sample as a function of height. Similar to the previous experiment, µ-SAXD radial and azimuthal maps were obtained, and the results are summarized in SI, Figure S3 and Figure 5, respectively, by the change in intensity in the azimuthal integration at q = 0.32 Å −1 (integrated from q = 0.29 Å −1 to q = 0.35 Å −1 ) along the scanning height. Here, the Sm-A orientation evolved spatially throughout the filament (see Video S2 and Figure 5), likely influenced by the gradient in stress from the printing process as well as the UV light penetration through the thickness of the filament. At the substrate interface (h~0 µm) the azimuthal angle intensity distribution was found to be narrow, suggesting strong alignment of the LCE layering. This region underwent a higher shear stress during printing compared to the bulk of the filament. As a consequence, the distribution of orientations increased, progressing in the bulk of the filament. The distribution can be quantitatively determined from 1D line cuts of the azimuthal intensity at various heights (see Figure S4). In the vicinity of the substrate interface, the layers were aligned parallel to the print direction, with an approximate distribution of ±2 • . As h increased, the distribution became wider to approximately ±3.5 • at a height of h = 50 µm and a distribution of ±10 • at h = 150 µm. Interestingly, besides the variation in the angular spread, the azimuthal angle distribution was maintained around 90 • , indicating a persistence of the Sm-A layering alignment from the bottom to the top of the filament. At the top of the filament (h between 210 and 240 µm), the intensity decreased, possibly due to the curvature of the filament at the top of the interface. It is also likely that local heterogeneities in the concentration of liquid crystal (which was 20 wt.%) greatly impacted the scattering intensity, producing the apparent streaks in intensity observed in the height scan.
Crystals 2021, 11, x FOR PEER REVIEW persistence of the Sm-A layering alignment from the bottom to the top of the filam the top of the filament (h between 210 and 240 μm), the intensity decreased, poss to the curvature of the filament at the top of the interface. It is also likely th heterogeneities in the concentration of liquid crystal (which was 20 wt.%) impacted the scattering intensity, producing the apparent streaks in intensity obs the height scan. In summary, after encountering initial external stimuli in situ during identified as compression in the barrel, shear stress during extrusion, and deform deposition pursued by relaxation and rearrangement post-deposition and during the LCE layering alignment followed the preferred direction induced by t direction. The specific alignment feature was then locked into the final struct preferential orientation of the Sm-A layering was particularly pronounced at the s interface. The preliminary investigation here demonstrated a technique mergi resolved μ-SAXD and 3D printing towards the control of the degree and directio layering orientation within the LCE-based printed scaffold. The μ-SAXD te provided a 2D mapping of the Sm-A orientation with layer normal appro perpendicular to the incident X-ray beam. We plan to conduct further ex situ rotating the sample with respect to the incident X-ray beam for a full 3D orientati In future work, we will explore the Sm-A layering alignment in rows of 3D filaments intending to fabricate a fully active and responsive cell support struct architectures to control cell directionality and anisotropy. This study will inv integration of a multitude of cell types into this elastomer-ink, to help it become a and not only provide but also stimulate cell alignment. Our goal is to fabricate other tissues, complex vascular networks from the LCEs using 3D-printing co cells. In summary, after encountering initial external stimuli in situ during printing identified as compression in the barrel, shear stress during extrusion, and deformation at deposition pursued by relaxation and rearrangement post-deposition and during curing, the LCE layering alignment followed the preferred direction induced by the print direction. The specific alignment feature was then locked into the final structure. The preferential orientation of the Sm-A layering was particularly pronounced at the substrate interface. The preliminary investigation here demonstrated a technique merging time-resolved µ-SAXD and 3D printing towards the control of the degree and direction of LC layering orientation within the LCE-based printed scaffold. The µ-SAXD technique provided a 2D mapping of the Sm-A orientation with layer normal approximately perpendicular to the incident X-ray beam. We plan to conduct further ex situ studies rotating the sample with respect to the incident X-ray beam for a full 3D orientation map. In future work, we will explore the Sm-A layering alignment in rows of 3D-printed filaments intending to fabricate a fully active and responsive cell support structure into architectures to control cell directionality and anisotropy. This study will involve the integration of a multitude of cell types into this elastomer-ink, to help it become a bio-ink, and not only provide but also stimulate cell alignment. Our goal is to fabricate, among other tissues, complex vascular networks from the LCEs using 3D-printing containing cells.