Volumetric 3D‐Printed Piezoelectric Polymer Films

A novel additive manufacturing route using a tailored resin containing Poly(vinylidene fluoride) Trifluoroethylene (PVDF‐TrFE) to 3D print piezoelectric films is demonstrated. Piezoelectric films are printed within 2 seconds in a single step by simultaneously focusing initiating and inhibiting excitations within the liquid resin to locally confine the photochemical reaction. The printed films are patterned with an array of holes with a diameter of 30 µm and a pitch of 55 µm. The piezoelectric response is homogeneous across the film, indicating that the print pattern does not impact the PVDF‐TrFE microstructure. Although the printed films contain only a small volume fraction of PVDF‐TrFE (3 wt.%), their piezoelectric response (d33 = 20.3 pC/N) is comparable to the highest literature values reported for PVDF‐TrFE films. The printed PVDF‐TrFE films are predominantly β‐phase, and no electrical poling, post‐processing, piezoelectric or inorganic additives are used in the fabrication. Analysis using piezoresponse force microscopy (PFM) and scanning electron microscopy (SEM) reveals that the enhanced piezoelectric response is due to the preferential formation of oriented PVDF‐TrFE phases during printing. These results demonstrate how the dedicated design of photoactive resins in combination with volumetric additive manufacturing can be applied to rapidly fabricate functional 3D structures.


Introduction
[3][4] Poly(vinylidene fluoride) (PVDF) is an organic polymer that can be solution-processed to form flexible films that exhibit ferroelectric, piezoelectric, and pyroelectric properties.PVDF offers many advantages in emerging applications as it does not contain metals or toxic elements, and is mechanically robust, flexible, lightweight, and even biocompatible. [5,6][8] The antiparallel packing of dipoles in the  and  PVDF phases leads to non-polar, insulating films, while the , , and  phases are polar and exhibit piezo-, pyro-, and ferroelectricity. [6,7]The phase is less thermodynamically favorable than the -phase but demonstrates the strongest piezoelectric response of all five phases.11][12][13] A challenge for PVDF piezoelectric applications is the lack of relevant fabrication protocols for industrial-scale production.[23][24][25][26] Further, hybrid fabrication approaches to increase the piezoelectric response include adding nanofillers to create compound materials including inorganic components, such as BaTiO 3 , PZT, Al(NO 3 ) 3 .9H 2 O, or graphene, clay, or biomolecules. [22,27,28]dditive manufacturing (AM) has recently emerged as an alternative fabrication route for low-cost prototyping and production of sophisticated structures that go beyond planar thin films.AM processes use either heat, light, and/or chemical reactions to convert adhesives into tailored geometries.Volumetric AM techniques, such as Xolography and Dual-Wavelength Volumetric Micro-Lithography (DWVML), rely on spatially-confined, optically-driven polymerization processes to fabricate 3D structures within the volume of a liquid resin. [29,30]Specifically DWVML is a single-photon process, that applies the excitation along the same pathway for both initiation and inhibition, enabling high-precision printing with fast throughput. [30]his is achieved by focusing a sequence of computed images along an optical axis.In contrast to other AM techniques that enable either high structural resolution on a small scale or rapid printing of larger structures, DWVML enables printing of micron-sized structures with objects 10 s of cm 2 large within seconds.
The main challenge of applying AM for PVDF-based piezoelectric applications is to develop photoactive PVDF-based resins that enable reliable, rapid printing of high-quality mechanical structures with excellent -phase yield.Photolithographic AM techniques pose further challenges, as these require the development of complex liquid resin compositions containing both photo-active macromolecules that chemically react to form a solid, such as PEGDA, [30,31] and chemically inert, photoinactive polymer chains of PVDF. [20]On one hand, the resin must be optimized with respect to the printing process, that is, the optical absorption and transmission, viscosity, and photochemical reactivity in the liquid.On the other hand, the resin must be optimized to enable a preferential microstructure for enhanced -phase yield of PVDF in the resulting solid print.
Here, we demonstrate a proof-of-concept for the 3D printing of piezoelectric PVDF-TrFE films via DWVML.The photochemical PVDF-TrFE-based liquid resin is converted into a piezoelectric 3D structured film within a single printing step that takes 2 s by focusing two excitation wavelengths, one to initiate the photochemical reaction and one to inhibit it, within the liquid resin.We optimize the photochemical resin with respect to the PVDF-TrFE content in order to facilitate the photochemical, volumetric AM printing of well-defined structures while achieving excellent piezoelectric response.No additional electrical poling or post-processing is used to enhance the piezoelectric properties of any of the samples in this study.We use piezoresponse force microscopy (PFM) to demonstrate that the piezoelectric response is homogeneous across the film, and not impacted by the printed structure.Despite the low PVDF-TrFE content (3 wt.%) within the non-piezoelectric resin matrix, the PVDF-TrFE-based prints demonstrate a piezoelectric response of 20.3 pC N −1 , which is comparable to reported values on films containing much larger piezoelectric material fractions. [9,12]We compare films that were fabricated photochemically by illuminating the PVDF-TrFEbased resins to PVDF-TrFE prints fabricated with DWVML and find that the piezoelectric response is higher and more homogeneous across the DWVLML print than in the photochemically fabricated films.These results indicate that, in addition to the resin composition, the fabrication process impacts PVDF-TrFE phase within the resin matrix.We apply scanning electron microscopy and Raman spectroscopy to show that the enhanced piezoelectric response in the printed films correlates with preferential PVDF-TrFE -phase formation and orientation within the resin matrix.

Structured PVDF-TrFE Films Printed via Dual-Wavelength Volumetric Micro-Lithography (DWVML)
Figure 1 shows the results of our optimized DWVML protocol for printing piezoelectric PVDF-TrFE films.The films were printed within seconds, and displayed excellent structural integrity, with well-defined structures.The printed 900 × 900 μm film is depicted in the photograph in Figure 1a on a € 0.10 coin for scale.Figure 2b shows an optical microscope image of the film, where the film structure (3 × 3 array of 300 μm cells) is visible.Each of the cells was printed in a single step that took 2 s.The cells consist of an array of holes with a diameter of 30 μm, and a pitch of 55 μm.
We chose this particular print structure in order to illustrate the potential to reliably print microstructured PVDF-TrFE prints using DWVML.On the one hand, it demonstrates how 3D microstructures (an array of circles with depressed surfaces, welldefined diameters, regular pitch, and sharp edges) can be reliably and reproducibly printed over the full sample area.On the other hand, the structure allowed us to reliably perform relevant characterization, such as PFM, Raman spectroscopy, and scanning electron microscopy, to determine the phase and microstructure of the PVDF-TrFE within the resin matrix.
To achieve these results, both the resin and the printing parameters were optimized to maximize the printing speed, as well as the structural resolution and structural integrity of the print.The liquid resin was first optimized with respect to light transmission, to facilitate the focusing of the initiating and inhibiting wavelengths within the volume of the resin, and viscosity of the resin, to print fine structures.Higher concentrations of PVDF-TrFE lowered the optical transmission while increasing the viscosity of the resin, and this in turn reduces the speed of photochemical conversion of liquid resin into solid film during printing and the resolution and quality of the prints.Therefore, we kept the PVDF-TrFE concentration low (with respect to the PEGDA concentration); in the optimized resin, PVDF-TrFE made up only 3 wt.% of the total solid resin.To maximize both the printing speed and resolution for our specific resin composition, the printing parameters were also optimized.To do this we investigated the influence of illumination intensity and duration of both the initiating (455 nm) and inhibiting (385 nm) excitations on the structure of the print (Figure 1c).The intensity of each excitation varied incrementally from 0 mW mm −2 to a maximum of 4.67 mW mm −2 for the initiating excitation and 3.75 mW mm −2 for the inhibiting excitation, respectively.To optimize the illumination conditions of the print, we use a custommade algorithm to detect contrast corresponding to solid structures.Red tones indicate full/high photo-conversion while blue areas indicate no/low conversion of liquid resin to solid film.Optimized printing protocols yield an abrupt difference and a visible cut-off between the region in the bottom right of Figure 1c) (high initiation excitation intensity/low inhibition excitation intensity) and the region top left of Figure 1c) (low initiation excitation intensity/high inhibition excitation intensity).This abrupt cut-off between liquid and solid indicates the formation of films with higher structural quality (sharp edges), whereas lower-quality  films display partially reacted regions (gel).Gel formation limits edge definition and occurs if there is an incomplete conversion of the liquid resin to solid print.Gel formation can be mitigated by increasing the optical transmission of the resin with respect to the initiating excitation, however, this is not practical for all resin compositions.It can also be achieved by increasing the intensity of the initiating excitation.Unwanted scattering as well as the extended duration of the initiating excitation, however, can limit the structural resolution of the print by converting some peripheral resin liquid to gel.To achieve well-defined structures, the intensity and duration of the inhibiting excitation must therefore be optimized with respect to the initiating excitation parameters.The regions marked with dashed circles in Figure 1c) depict their polymerization rates as a function of time in Figure 1d) and are used as an optimization route to obtain high-quality prints.In region I, we applied intermediate intensity for the initiation excitation combined with low intensity for the inhibition excitation, resulting in polymerization and high conversion of liquid resin to film.In region II, the inhibition excitation was increased with respect to I, resulting in suppression of the photochemical reac-tion.Finally, in Region III, we decrease the intensity of the initiation excitation, while maintaining the high intensity of inhibition excitation, thereby completely suppressing polymerization.

Impact of the Resin Matrix on the Piezoelectric Response of PVDF-TrFE Films
The total PVDF-TrFE content in the resin was 3 wt.%,while the PEGDA oligomer and its initiator/co-initiator pair and inhibitor made up ≈97 wt.% of the solid components.The photopolymerization of PEGDA creates the matrix of the solid print, and we expect this to impact both the microstructure of the PVDF-TrFE as well as the piezoelectric response of the print.To investigate this, we performed PFM measurements on DWVML PVDF-TrFE films (Figures 2 and 3) and compared the results to photochemically (PC) fabricated PVDF-TrFE films.Both samples were prepared using the same resin.The difference between these samples is that DWVML films are rapidly printed using both initiating and inhibiting excitations to obtain fine structures, while the PC films were produced by exposing the resin only to the initiating excitation until the resin had completely solidified (see Experimental Section), resulting in a thick, planar film.
Figure 2 depicts AFM and PFM maps taken from the same regions on the PC and DWVLM films.We observe that increasing PVDF-TrFE content is correlated with increasing roughness in the AFM of the PC films.It is worth observing that while the AFM reveals some topographical variance in the PVDF-TrFE films, no inhomogeneities are observed in the PFM phase maps.Therefore, we postulate that morphological changes on this length scale (μm) do not correlate to variations in the piezoelectric response.
We compare the PFM results from the PVDF-TrFE films to PFM results taken on an inorganic piezoelectric reference sample (aluminum scandium nitride, Al 1-x Sc x N) measured under the same conditions.An amplitude of 1.4 mV measured in our DWVML samples would correspond to a piezoelectric coefficient (d 33 ) of ≈20.3 pC N −1 .,32] The piezoelectric response of the PC samples is considerably lower at 12 pC N −1 , yet still comparable to previously published values.In both cases, we note that the piezoelectric response is remarkable considering the low solid content of PVDF-TrFE in the films (3 wt.%).This indicates that the photochemical polymerization of the resin may promote the formation of preferentially aligned -phase PVDF-TrFE domains in the samples.To confirm that the resin impacts the microstructure of PVDF-TrFE in the films, we prepared resins with different relative concentrations of PVDF-TrFE to PEGDA (ranging from 3 to 16 wt.%,see Table S1 and Figure S2, Supporting Information).
Figure 3 depicts the PFM amplitude and phase of DWVML (3 wt.%) and PC (16, 11, 7, and 3 wt.%)PVDF-TrFE films.Both fabrication methods resulted in single-polarity piezoelectric PVDF-TrFE, however, we observe that the PFM response is higher as well as more homogeneous in the DWVML film than in the PC films.This can be observed by the narrow amplitude and phase distributions in Figure 3a,b, respectively, for PC compared to DWVML films.The PC films have the highest mean amplitude centered at ≈0.8 mV, while the DWVML film yields a mean PFM amplitude, centered at ≈1.4 mV.We note that the DWVML fabrication route leads both to an increase in PFM amplitude, as well as more homogeneous films than the PC fabrication route.Interestingly, as the PVDF-TrFE content decreases in the PC films, we observe a general increase in piezoelectric response, although we cannot identify a clear trend.

Impact of Resin Matrix on PVDF-TrFE Phase and Chain Conformation
We apply Raman spectroscopy to investigate the impact of the photochemical reaction on the formation of -phase PVDF-TrFE, as well as on-chain conformation.To do this we compare the Raman spectra of powder PVDF-TrFE and spin cast PVDF-TrFE films with PVDF-TrFE prepared using PC and DWVML.The Raman spectrum of PVDF-TrFE, which has been documented in detail in previous literature, [22,23,[33][34][35][36][37][38] can be seen in Figure 4a) for PVDF-TrFE powder (red), SC PVDF-TrFE (green), PC PVDF-TrFE (purple), and DWVML PVDF-TrFE (black, dashed).
The Raman peak centered ≈800 cm −1 corresponds to the Gauche conformation of the carbon backbone that is associated with the -phase of PVDF-TrFE.The Raman peak centered ≈840 cm −1 corresponds to the symmetric stretching of the CF 2 bond, attributed in the literature to the -phase of PVDF-TrFE. [39]he peak centered ≈880 cm −1 is associated with the CH 2 rocking vibration and CF 2 asymmetric stretching in PVDF-TrFE, and is found in both  and -phase. [39]We exploit this to quantify and compare the relative -phase content in the different PVDF-TrFE samples.We note that the very thin DWVML PVDF-TrFE film (Figure 4a), dashed line) did not yield a sufficiently strong Raman signal for quantitative analysis.We, therefore, focus our analysis on the powder, spin cast, and PC PVDF-TrFE samples, including PC films prepared using different PVDF-TrFE concentrations (Figure 4b).The DWVML films were prepared with resins containing 3 wt.%PVDF-TrFE, for reference.The Raman spectra of the individual resin components (EDAB, o-Cl-HABI, CQ & PEGDA) are shown in Figure S1 (Supporting Information).
We compare the relative -phase in the PVDF-TrFE samples with the ratio in Raman peak intensity at 840 cm −1 (I 840 ; -phase) and 800 cm −1 (I 800 ; -phase), that is / = I 840 /I 800 .All samples yielded /-ratios > 1, and SC PVDF-TrFE shows the lowest value (1.48), followed by PC PVDF-TrFE (2.87), which reveals a /-ratio slightly below powder PVDF-TrFE (2.96).We note that the Raman spectra of PEGDA also reveal a peak centered ≈800 cm −1 (Figure S1, Supporting Information), and therefore / values extracted here may underestimate the actual values, especially in films with lower PVDF-TrFE concentrations.However, the Raman spectra reveal that the PVDF-TrFE in the films is predominantly -phase, which is further corroborated by X-ray diffraction (XRD) measurements that revealed diffraction peaks corresponding to lattice parameters of the polar orthorhombic phase as published by Hasegawa et al. [40] (Figure S3, Supporting Information).The correlation between Raman peak center and stress/strain on polymer backbone has been previously reported for PVDF nanocomposites and other crystalline polymers. [39,41]A redshift of the Raman peak center shift is associated with polymer backbone elongation and higher strain, while a blueshift is associated with compressive stress. [41]The change in the Raman mode can be seen primarily at vibrational modes associated with the skeletal CC bond stretching (at 610 cm −1 ) and CCC bond angle deformation (at 880 cm −1 ). [41]In our samples, we clearly observe red shifting in the Raman modes of the PC and AM PVDF-TrFE film (Figure 4a), which we attribute to the elongation of the PVDF-TrFE chains within the resin matrix.This is a further indication of increased -phase formation, as computational simulations have shown longer carbon backbone bond lengths in -phase PVDF. [42]

Impact of Resin Matrix on the Microstructure of PVDF-TrFE Films
To investigate the PVDF-TrFE microstructure in the resin matrix in more detail, we used field-emission scanning electron microscopy (FE-SEM) to study the DWVML and PC films.We exploit the contrast in backscattered electrons based on the atomic number between PVDF-TrFE and PEGDA to investigate the microstructure and phase distribution; the PVDF-TrFE appears bright within the darker PEGDA matrix.We observe that PVDF-TrFE crystallites in both the DWVML film and PC films are distributed uniformly over the film surface as well as throughout the thickness of the sample.In the DWVML film, the crystallites are smaller, uniform rod-like structures (Figure 5a) while in the PC film, the crystallites are larger, web-like structures (Figure 5c).The cross-section image of the DWVML film demonstrates rod-like crystallites throughout the thickness of the sample (Figure 5b).The rod-like crystallites appear to align in parallel, while in contrast, the structures in the PC film appear to be isotopically dispersed within the matrix.Darker vertical lines visible here are attributed to the curtain effect, an artifact of the ion beam preparation used for the cross-section measurements.
Very low concentrations of PVDF-TrFE (3 wt.%) were used to achieve the required optical properties and viscosity to enable the rapid printing (2s) of high-quality micro-structured prints.Here we show that the low content of active piezoelectric material, however, is compensated by preferential alignment of the PVDF-TrFE chains within the resin matrix.The Raman spectra reveal that PVDF-TrFE preferentially forms solids in the -phase (Figure 4a).In the case of PC and DWVML films, we additionally observe red shifting of Raman modes associated with backbone elongation and compression at 610 and 880 cm −1 .These findings correlate strongly with DFT simulations done on and -phase PVDF. [42]While the PC and DWVML films were prepared with the same resin (Figure 5), we observe that the PVDF-TrFE phases in the DWVML film form smaller, rod-like clusters, and appear anisotropically ordered with preferential parallel alignment of the PVDF-TrFE phase with respect to the substrate.This results in an enhanced piezoelectric response in DWVML films compared to PC films.These results demonstrate that not only the properties of the solid resin matrix, but also the photochemical reaction dynamics, can be used to tailor the piezoelectric properties of the print.

Summary
We demonstrated a proof-of-concept for 3D printing of PVDF-TrFE films with a high piezoelectric coefficient (d 33 = 20.3pC N −1 ).The films were fabricated using a singlephoton, two-wavelength, volumetric printing route (DWVML) from a liquid resin in a single step in 2 s.The 900 × 900 μm printed films contain an ordered array of holes with a diameter of 30 μm and a pitch of 55 μm.Despite the microstructured texture, we measure a homogeneous piezoelectric response across the print.The prints contain very low PVDF-TrFE content (3 wt.%), however the piezoelectric response is comparable to literature values of PVDF-TrFE.Further, printed PVDF-TrFE films demonstrated higher piezoelectric coefficients than samples prepared from the same resin using a simple photochemical route.We employ AFM, Raman spectroscopy, and FE-SEM to establish a correlation between the enhanced PFM signals observed in printed films and the preferential distribution of the PVDF-TrFE phase, as well as the underlying microstructural features within the resin matrix.We found that the piezoelectric response is anti-correlated with the PVDF-TrFE content and surface roughness.In addition to PVDF-TrFE concentration in the resin, the fabrication route has a significant impact on the piezoelectric response.DWVML films exhibited not only better piezoelectric characteristics than PC films but also demonstrated less variability in the PFM response across the film.We attribute the enhanced piezoelectric response observed here to preferential PVDF-TrFE phase formation in the resin.In the case of DWVML films, PVDF-TrFE microstructures were observed to anisotropically align preferentially with respect to the substrate.These results demonstrate how volumetric printing can be applied to control resin morphology, including molecular arrangement, orientation, and crystalline phase formation to enable the rapid printing of microstructured films with superior piezoelectric properties, even with low concentrations of active material.The resin used here was based on a recipe developed by Photosynthetic B.V. [30] that exploits the photopolymerization of the PEGDA oligomer and contains PEGDA, o-Cl-HABI (photoinhibitor), CQ, and EDAB (initiator and co-initiator pair) dissolved in THF.The PEGDA-based resin was then blended with PVDF-TrFE, giving a solid polymer content of 3.07% PVDF-TrFE.

Experimental Section
All of the materials were stored in dry, dark, and cool conditions or otherwise according to the recommended storage conditions.

Deposition of PVDF-TrFE Films-3D Printing of PVDF-TrFE-Based Resin:
The resin(50 μL) was used for a single print.The setup contains two highprecision LED excitation sources.The initiating 455 nm (4.67 mW mm −2 ) and inhibiting 385 nm (3.75 mW mm −2 ) excitations were emitted by an Agilent Tech LED focused within the volume of the resin, where vertical movement of the focal point is controlled via stage translation.Python code was used to optimize the illumination duration and area to achieve welldefined 3D structures, with a resolution of ≈1 μm over an illumination area of 300 × 300 μm into a flat printing volume of ≈1 mm × 50 mm × 50 mm.The unreacted liquid resin was subsequently flushed from the sample, and the structure was then cured for 30 s in blue light (450-470 nm).No postprocessing steps, such as poling, were applied to the films.

PVDF-TrFE Films Produced via Photochemical Reaction:
To optimize the composition and photochemical properties of the resin, A simplified photochemical route was used to produce larger planar (25 × 25 mm) PVDF-TrFE films.This larger sample geometry facilitated structural characterization via XRD and Raman spectroscopy.To fabricate these samples, liquid resin was doctor-bladed on the substrate (200 μm).The resin was subsequently irradiated under constant illumination for 1 h until the resin solidified into a film.A 100 W Oriel mercury lamp was used in conjunction with an Edmund Optics TECHSPEC 400 and 50 mm diameter, OD 2.0 Fused Silica (Corning 7980) longpass filter.Under these conditions, the photochemical reaction could proceed unimpeded, as the filter blocked lower energy wavelengths that inhibit the reaction.No post-processing steps, such as poling, were applied to the films.
Spin-Casting PVDF-TrFE: Spin-cast PVDF-TrFE samples were produced as a reference for Raman spectroscopy studies.A solution of Ace-tone and N-N dimethylformamide (DMF) in a 4:1 ratio was mixed with PVDF-TrFE at 30% w/w under sonication for 1-3 h.The spin-coater was operated at ambient conditions with the lid closed at 3000 RPM for 45 s and followed by 6000 RPM for another 2.5 min.The resulting film was then annealed at 100 °C for 1 h.No post-processing steps, such as poling, were applied to the films.
Characterization of PVDF-TrFE-Piezoresponse Force Microscopy (PFM): The piezoelectric properties of PC and DWVML PVDF-TrFE films were analyzed by PFM with the atomic force microscope (NX-20, Park Systems).The PFM measurements were carried out in contact mode with a PtIr5coated silicon tip (PPP-EFM, NanoSensors) with a resonance frequency of ≈75 kHz.PVDF-TrFE films were fabricated on pre-cleaned ITO-coated glass substrates (Präzisions Glas & Optik GmbH; ≤ 10 Ohm sq −1 ), and the ITO served as the bottom electrode while the tip served as the top electrode.
During the PFM measurements, an AC voltage with 10 V amplitude and frequency of 17 kHz was applied through the AFM tip (off-resonance PFM mode). [43]The deflection of the tip caused by the oscillations of the sample surface (inverse piezoelectric effect) was measured with a lock-in amplifier.

Characterization of PVDF-TrFE-High-Resolution Field Emission Scanning Electron Microscopy (FE-SEM):
The surface and cross-section of PVDF-TrFE films were imaged with high-resolution field emission scanning electron microscopy (FE-SEM) (SU8000, Hitachi).Cross-sections of the films were prepared by a broad ion beam technique (BIB) with the cross-section polisher (SM-09010, Jeol).For the preparation, an argon ion beam with 6 kV accelerating voltage and a full width of half a maximum of 500 μm was used.A rocking of the sample ± 45°during the ion milling was applied to reduce the curtaining effect.
Characterization of PVDF-TrFE-Raman Spectroscopy: Raman spectra were obtained with the Renishaw inVia Raman Microscope, equipped with a 785 nm laser and a ×50 Leica optics objective.Five integrations with an exposure time of 10 s were combined to obtain a single spectrum.A set of 10 Raman spectra was taken for each sample measurement to confirm that the samples did not degrade during measurement, as well as confirm that the data does not contain artifacts due to laser heating, focus shift, or laser instability.Samples were measured on either glass or quartz substrates without encapsulation.

Figure 1 .
Figure 1.A DWVML printed PVDF-TrFE film is shown in (a) photograph with a € 0.10 coin for scale, and in (b) optical microscope image at 4× magnification.The printed film consists of 9 cells, each of which contains an array of holes with a diameter of 30 μm and a pitch of 55 μm.The test print for optimizing the photochemical reaction for the DWVML protocols is shown in Figure (c) and (d).Figure (c) shows the variation in illumination intensity of the initiating excitation (x-axis) and the inhibiting excitation (y-axis).The solid printed areas are represented by brighter colors, while darker colors represent unreacted resin.Three distinct illumination conditions are marked in regions I, II, and III. Figure (d) depicts the conversion of liquid resin to a solid film over 6 s for the illumination conditions I (intermediate inhibiting light & low initiating light intensity), II (intermediate inhibiting light & high initiating light intensity), and III (low inhibiting light & high initiating light intensity).
Figure 1.A DWVML printed PVDF-TrFE film is shown in (a) photograph with a € 0.10 coin for scale, and in (b) optical microscope image at 4× magnification.The printed film consists of 9 cells, each of which contains an array of holes with a diameter of 30 μm and a pitch of 55 μm.The test print for optimizing the photochemical reaction for the DWVML protocols is shown in Figure (c) and (d).Figure (c) shows the variation in illumination intensity of the initiating excitation (x-axis) and the inhibiting excitation (y-axis).The solid printed areas are represented by brighter colors, while darker colors represent unreacted resin.Three distinct illumination conditions are marked in regions I, II, and III. Figure (d) depicts the conversion of liquid resin to a solid film over 6 s for the illumination conditions I (intermediate inhibiting light & low initiating light intensity), II (intermediate inhibiting light & high initiating light intensity), and III (low inhibiting light & high initiating light intensity).

Figure 2 .
Figure 2. Juxtaposition of AFM topography and PFM phase maps from identical regions of the respective films.The PC films were fabricated from resins containing a) 16 wt.%,b) 7 wt.%, and c) 3 wt.%PVDF-TrFE, and the DWVML film was printed from a resin containing d) 3 wt.%PVDF-TrFE.
Figure 5a,b shows the surface and cross-section SEM images of the DWVML PVDF-TrFE film, and for comparison, Figure 5c,d shows the surface and cross-section images of a PC PVDF-TrFE film prepared from the same resin (3 wt.% PVDF-TrFE).

Figure 5 .
Figure 5. FE-SEM images of DWVML PVDF-TrFE film a) surface and b) cross-section and of the PC PVDF-TrFE film c) surface and d) cross-section.