3D-Printed Piezoelectric Stents for Electricity Generation Driven by Pressure Fluctuation

Abstract

Vascular stenting is a commonly used procedure to support narrowed blood vessels by opening the passageway and restoring blood flow to ischemic tissues in the heart and other organs. However, in stent stenosis resulted from the vascular inflammatory and hyperproliferative responses triggered by stent deployment restricts blood flow and may cause subsequent life-threatening issues. Drug coating is a frequent strategy to limit stenosis and prolong stent patency. In this study, we sought to develop a novel class of stent utilizing fused deposition modeling 3D printing with built-in electric field to construct a piezoelectric stent featuring a zigzag shape. The biocompatible piezoelectric composite was made of potassium sodium niobite microparticles and poly (vinylidene fluoride-co-hexafluoropropylene), complementing each other with good piezoelectric performance and mechanical resilience. The in-situ poling yielded an appreciable piezoelectricity (d₃₃ ~ 4.2 pC N⁻¹) of the as-printed stents. Under stimulated blood pressure fluctuation, the as-printed piezoelectric stent was able to generate peak-to-peak voltage from 0.07 V to 0.15 V in corresponding to pressure changes from 20 Psi to 120 Psi, giving a sensitivity of 7.02E-4 V Psi⁻¹. Biocompatible piezoelectric stents bring potential opportunities for real-time monitoring blood vessels or enabling therapeutic functions.

Graphical Abstract

1. Introduction

Stenting is a common strategy to treat peripheral arterial disease,¹ colonic strictures,² ureteral obstructions³ and coronary artery narrows.⁴ Stents are typically minimum-invasively delivered to the lumen of obstructed vessels via catheters. Based on the mechanism of expanding, stents are designed to be self-expandable or balloon-expanded. After deployment, stents should be kept on site over their entire lifetime, enduring the dynamic blood flow and pressure fluctuations. This application environment places stent at a unique position to constantly interface with biomechanical agitations. Most current medical stents are made from ultrafine stainless steel wires weaved into a zigzag network, which provides mechanical support to widen and hold the lumen of anatomic vessels to maintain the circulation of blood. If utilized effectively, these continuous biomechanical movements may provide an endless energy source to support local and in vivo electronic functionalities, such as electrical stimulation, drug delivery, tissue healing or anti-thrombosis, which are the future of smart implantable devices.⁵

Recent advancements in biocompatible piezoelectric materials enabled in vivo coupling between biomechanical motions and localized electricity generation in biological systems.⁶ The incorporation of biocompatible piezoelectric materials arouses the revolution of biomedical devices, offering unprecedented functionalities for energy harvesting,^{7, 8} drug delivery,⁹ actuation¹⁰ and physiological sensing.¹¹ We envision that stents made of biocompatible piezoelectric materials may be capable of self-sustainable in vivo electricity generation,¹² leading toward multi-functionality in future stents technology. Our prior research demonstrated a capability of fabricating piezoelectric artificial blood vessels by electric field-assisted 3D printing.¹³ The in situ poling strategy allowed direct printing of complex 3D lattices from piezoelectric composites.¹⁴ Leveraging this technique, in this work, we designed and fabricated piezoelectric stents using fused deposition modeling (FDM) 3D printing with in situ poling. The stents are designed with a zigzag shape made from composite wires of potassium sodium niobate (KNN) particles and poly (vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP). The selection of the zigzag shape is expected to facilitate stent placement and 3D printing process due to its proven flexibility and feasibility. The excellent combination of piezoelectric and mechanical properties in this 3D printable ferroelectric composite has been validated in our previous work, where KNN particles make primary contributions to the piezoelectric responses as well as improve the mechanical strength in a flexible PVDF-HFP matrix.¹⁵ The stents showed desired mechanical properties and piezoelectric property. Therefore, they can produce appreciable piezoelectricity when subject to simulated mechanical strains, demonstrating a possibility for in vivo sensing¹⁶ and electricity generation driven by regular blood pressure fluctuation.¹⁷

2. Experimental Sections

Preparation of piezoelectric composite: The piezoelectric composite was made from ferroelectric K_0.5Na_0.5NbO₃ (KNN) and PVDF-HFP. KNN microparticles was prepared by standard solid-state reaction method, consisting of potassium carbonate (K₂CO₃, > 99%, Sigma), sodium carbonate (Na₂CO₃, > 99.5%, Sigma) and niobium (Ⅴ) oxide (Nb₂O₅, > 99.9%, Chemsavers) at a molar ratio of 1:1:2. The pile of K₂CO₃, Na₂CO₃ and Nb₂O₅ powder was mixed in a nylon mill jar including zirconia grinding balls and further grinded together with ethanol for 2 h by a gear-drive planetary ball mill machine (BM4X-V2.0L, Col-Int Tech, LLC). Afterward, the soupy mixture was dried at 60 °С in oven overnight to dehydrate. Large lumps were removed by using a 100-mesh lab sieve from the fully dry mixture, and then crushed and grinded by mortar and pestle sets for 2 h. The subtle particles were sintered in a muffle furnace at 1100 °C for 4 h and then cooled down naturally to room temperature. Ferroelectric KNN microparticles were achieved after grinding manually for another 2 h. PVDF-HFP pellets diced from Fluorinar-C^™ PVDF filament (NPFC175W1000, Nile Polymers, LLC) were dissolved in N,N-dimethylformamide (DMF, > 99.8%, Sigma) solvent with 20 wt% solute, followed by stirring for 1 h at 80 °С. Corresponding prepared KNN microparticles with 139 wt% of PVDF-HFP were added into the uniform solution and kept stirring until completely mixed. The final mixture was transferred into fume hood to evaporate DMF solvent and then kept dry in oven at 70 °С overnight, subsequently grinded within liquid nitrogen by a multifunctional electric grinder (GS-810N, Secura, LLC) into fine piezoelectric composite.

Extrusion of piezoelectric filament: The as-prepared KNN/PVDF-HFP composite was added into a polymer extruder (assembled from Filastruder kit) and extruded through a 2.85-mm nozzle at 162 °C. The size of the extruded piezoelectric filament was assessed and controlled by a filament guide with a diameter of 2.85 mm.

3D Printing of piezoelectric stents: The piezoelectric stent was printed by a custom fused deposition modeling 3D printer (LulzBot TAZ Pro 3D Printer, Lulzbot, LLC) through a 0.5-mm printing nozzle at 240 °C. A rotating stainless-steel rod with a diameter of 6 mm was driven by a 12 V gear motor and used as the supporting substrate. The surface of the rod was coated with a thin layer of PVA glue (Elmer’s Disappearing Purple Glue Stick, Elmer’s, LLC) to improve the adhesion of printed materials. The head nozzle kept printing and moving back-forward along axial direction at 2.5 mm per second, meanwhile the supporting rod was rotating at 0.2 revolves per second. With the simultaneous motion of printing nozzle and rotating rod, a zig-zag shaped piezoelectric stent including five rings connected by an inclined bridge was printed out. The printing nozzle and rotating stainless-steel rod were connected respectively to the positive and negative electrodes of a high voltage source (30 kV, 10 W) during the entire printing process. The piezoelectric stent with 6 mm internal diameter was polled simultaneously with a built-in electric field of 1.2 kV mm⁻¹ during 3D printing process.

Morphology, structure and piezoelectric characterization: The surface morphology was captured by a fluorescence LED microscope (Leica DM4 M) and a Zeiss Gemini 450 field-emission scanning electron microscope (SEM). X-ray diffraction (XRD) patterns of piezoelectric composite was characterized by a Bruker D8 Discovery X-ray powder diffractometer with 2-d detector. Ferroelectric property of piezoelectric composite, represented by a square layer (0.5 cm × 0.5 cm) of printed piezoelectric thin film, was measured by a polarization–electric field (P–E) hysteresis loop conducted by a P-E loop workstation (Premier II, Radiant Technologies Inc., Albuquerque, NM, USA). PFM amplitude and phase characterizations were performed on a printed thread with 0.5 mm in diameter using an XE-70 Park System. The piezoelectric coefficients d₃₃ of printed thread with 0.5 mm in diameter were obtained by a quasistatic d₃₃ piezometer (ZJ-3A, Institute of Acoustics, China). All compressive and tensile modulus tests of printed stents including cyclic compression-release were measured by a Rheometrics Solids Analyzer III dynamic mechanical analyzer.

Evaluation of Biocompatibility: Cell Morphology Examination. UV-sterilized samples of the KNN/PVDF-HFP were placed on the inside of the transwell insert of a 24-well plate with clear walls and bottom (Corning 3422). No material was added to the control wells. There were four wells in each group. In total, 1.66 × 103 MOVAS cells were seeded into each well containing 700 μL of cell culture medium. After 24, 48 and 72 h, the cytoskeleton and nucleus of the cells were stained with Alexa Fluor 555 Phalloidin (Thermo Fisher Scientific, A34055) and DAPI, respectively. After staining, the cells were imaged using a Keyence BZ-X800 fluorescence microscope.

Biocompatibility Evaluation. MOVAS cells were purchased from American Type Culture Collection (ATCC, CRL-2797) and grown as recommended in modified DMEM containing 4.5 g L⁻¹ D-glucose (Thermo Scientific, 11965118) supplemented with 10% fetal bovine serum (FBS), 100 U mL⁻¹ penicillin, and 100 U mL⁻¹ streptomycin. The cell biocompatibility of the KNN/PVDF-HFP was assessed with a CellTiter-Glo assay (Promega, G9242) as described. UV-sterilized samples of the KNN/PVDF-HFP were placed inside of the transwell inserts of a 24-well plate with clear walls and bottom (Corning 3422). No material was added to the control wells. There were three wells in each group. In total, 1.66 × 104 cells were seeded into each well containing 700 μL of cell culture medium. After incubation for 24, 48 and 72 h at 37 °C and 5% CO₂, the cell culture medium was aspirated and 350 μL of PBS and 350 μL of CellTiter-Glo solution were added to each well followed by incubation at room temperature for 30 min. Luminescence was recorded on a Varioskan LUX microplate reader (Thermo Scientific). The relative cell viability was expressed as (luminescence of sample wellsblank) (luminescence of control wells-blank) × 100%, where blank is the luminescence of the wells without cells (PBS and CellTiter-Glo solution only).

Piezoelectric output measurements: The printed stents were settled tightly on a gold-coated polydimethylsiloxane (PDMS) hollow tube with an external diameter of 6 mm. Another layer of PDMS was applied to fill the surficial space to avoid short circuit, followed by a second layer of gold (50 nm in thickness) using e-beam evaporation. Two gold threads were subsequently attached to the first and second layer of gold electrode, respectively. A balloon system (Cordis, Corporation, USA) replete with flowing fluids was inserted into the PDMS hollow tube to puff up the whole device. A syringe pump affiliated with the balloon was loaded with a pressure gauge ranging from 0 to 440 Psi to control and monitor the internal fluid pressure. At the same time, the syringe pump was driven by the up-and-down motion of a computer-controlled actuator (LinMot) to 5 mm, 10 mm, 12 mm and 15 mm, corresponding to 20 Psi, 70 Psi, 90 Psi and 120 Psi. The piezoelectric voltage output of the fabricated device was recorded by a multimeter (DMM 6500, Keithley) and filtered by Band block FFT off 59.9–60.1 Hz (Figure S1). The voltage measurement during the long-term stability test of the piezoelectric performance was performed by an electrometer (6515 System, Keithley).

3. Results and Discussions

The piezoelectric stent was designed with a zig-zag shape, similar to one of the common structures of clinically used stents. This stent structure was directly printed on a rotating stainless-steel rod, as schematically shown in Figure 1a (Supporting video S1). The rod was used as a supporting surface for 3D printing, as well as for providing a shape template to define the size of the stent. Here, a rod with a diameter of 6 mm was used, which could produce ~6.5 mm stents that fit a 6.5 mm sheath. A high voltage of 1.2 kV was applied between the printer nozzle and the stainless-steel rod during the entire printing process. It introduced an in-situ poling electric field of 1.2 kV through the as-printed ferroelectric filament to align its polarization. The rotation rate (1 rotation per 5 seconds) was coordinated with the nozzle movement speed and position to create a zigzagged ring pattern continuously along the axial direction (see experimental section for detail and Figure S2). Through this approach, a stent made of piezoelectric filament (Figure S3a) could be directly printed at any selected length with a uniform diameter. Bottom inset of Figure 1b shows an as-printed stent with an internal diameter of 6 mm comprising 5 zigzag rings giving a total length of 2.4 cm. The zigzag lattice allowed the stent to be compressed into smaller size to fit into a catheter and released by regular balloon expansion (Figure S4, Supporting video S2).

Figure 1.

Piezoelectric composite stents. (a) Schematic illustration of the stent fabrication setup by 3D printing with in situ poling. Inset: a digital photo of 3D printing set-up. (b) Schematic illustration of the piezoelectric stent structure. Inset is a digital stent photo of an as-printed stent. (c) Optical picture of the as-printed stent surface. (d) SEM image of the as-printed stent surface showing a uniform melted mixture of KNN/PVDF-HFP. (e) Higher resolution SEM image showing cubic KNN microparticles imbedded in the PVDF-HFP matrix.

The strong in situ poling electric field could align the randomly oriented KNN ferroelectric domains in the stent filament along its radial direction (top inset of Figure 1b).¹⁸ Therefore, the radial direction was considered as the out-of-plane direction, and the generated piezoelectric potential was defined between the outer and inner surfaces of the stent (Figure 1b). As reported in our previous work,¹³ the large d₃₃ value and high curie temperature (~440 °С) of KNN allowed the piezoelectric phase to remain stable at the printing temperature (240 °С). PVDF-HFP served as a flexible and thermostable polymer matrix that brought a high structural integrity to the stent.¹⁹ Because the printing temperature was much higher than the curie temperature of PVDF-HFP (~120 °С), majority PVDF-HFP was in a non-ferroelectric phase. However, our previous observed a small portion of the ferroelectric β-phase could also form driven by the polarized ceramic-polymer interface,²⁰ which may introduce additional contribution to the piezoelectricity of the printed structure.²¹ The fluorine-based polymer could also offer a hydrophobicity surface that is desired for blood interaction.²² KNN microparticles combined with PVDF-HFP matrix, forming a versatile piezoelectric stent with line-scale d₃₃ over 4.2 pC N⁻¹ (Figure S5). This value is in the comparable range of other piezoelectric polymers, such as PVA/PVDF hydrogel composite (0.9 to 8.4 pC N⁻¹)²³ and PLLA fibers (7 to 12 pC N⁻¹),²⁴ and could be further improved by raising the poling electric field and adjusting the composition ratio.¹³

The morphology of as-printed stent filament was first examined under an optical microscope. As shown in Figure 1c, the stent filament had a uniform thickness of ~250 μm printed through a 0.25-mm nozzle. The stent surface was smooth, and no abrupt surface extrusion or defects could be observed. The cloudy gray dots throughout the filament body indicated a uniform distribution of the KNN microparticles in the PVDF-HFP matrix. The uniform mixture of the KNN/PVDF-HFP composite was further visualized by scanning electron microscopy (SEM) image (Figure 1d). The as-printed surface had a rather low surface profile and did not show any feature of particle agglomerations. Higher resolution SEM image clearly revealed well-proportioned distribution of largely imbedded KNN microparticles with a cubic shape and a size of ~ 1–2 μm (Figure 1e). This homogeneous mixture is important to ensure uniform piezoelectric responses and good mechanical integrity.

X-ray diffraction (XRD) characterization was then used to confirm the presence of the ferroelectric phases in the as-printed composite filament. The XRD spectrum revealed two sets of diffraction peaks (Figure 2a). The strong 2θ peaks near 22.8°, 31.8°, 46.9° and 52.2°, according to JCPDS card 77–0038, were from the ferroelectric KNN microparticles.²⁵ One more small peak at 20.8° was also observed corresponding to the β phase of PVDF-HFP,²⁶ suggesting that a small portion of PVDF-HFP was crystalized into its ferroelectric phase possibly due to the interface influence. The lower peak intensity of PVDF-HFP compared to KNN microparticles in XRD patterns was due to its lower degree of crystallinity. The ferroelectric property of the as-printed composite (in a thin film morphology) was evaluated by polarization-electric field (P-E) hysteresis loop (Figure 2b). Measured from a 0.5-mm thick film, the polarization could reach a saturation point as high as ~1.75 μC cm⁻² at an electric field of ~6 kV mm⁻¹. The remnant polarization at zero electric field also had an appreciable value of ~0.56 μC cm⁻², suggesting it could serve as a good piezoelectric build block for electromechanical coupling.

Figures 2.

Piezoelectric property of as-printed stents. (a) XRD characterization of KNN/PVDF-HFP composite powder. (b) Polarization–electric field (P–E) hysteresis loop of 3D printed ferroelectric composite film at room temperature. (c) AFM topography images overlaid with corresponding PFM amplitude contrast (i, iv), and PFM amplitude (ii, v) and phase images (iii, vi) of as-printed KNN/PVDF-HFP samples with polling (i, ii, iii) and without polling (iv, v, vi).

The localized piezoelectric property was mapped from the as-printed piezoelectric filament surface by piezoresponse force microscopy (PFM) to reveal the distribution of surface piezoelectric polarization at the sub-micrometer scale (Figure 2c). PFM characterizations were conducted on samples with (i, ii, iii) and without (iv, v, vi) applying the electric field during printing to confirm the effect of in situ poling. First, the surface topographic images revealed a vertical morphology variation of ~200 nm from the poled sample (Figure 2c-i), while the un-poled surface exhibited a surface fluctuation of ~400 nm (Figure 2c-iv). The higher surface smoothness from the poled samples suggested that the in situ electric field could be helpful for driving the KNN microparticles deeper into the polymer matrix. Figure 2c ii and iii revealed the amplitude and phase of the inversed piezoelectric response of the polled sample. Here, the amplitude indicated the magnitude of surface deformation due to the applied electric field at the tip-sample interface, while the phase indicated surface dipole orientation. From the poled configuration, we observed a strong amplitude displacement shown by the bright intensities (Figure 2c-ii) approaching ~100 mV. Furthermore, the phase analysis showed a uniform phase contrast which indicated a uniform dipole alignment of ~180° across the surface of the poled sample (Figure 2c-iii). In contrast, from the un-poled surfaces, the amplitude responses only reached a magnitude of ~80 mV (Figure 2c-v) and the phase contrast was less uniform across the surface (Figure 2c-vi). Overlaying the amplitude images with the corresponding surface topography images did not show a strong polarization and morphology correlation, suggesting the minimal topological influences to the PFM measurements. These measurements together highlighted the importance of the in situ applied electric field during the printing process.

As the essential function of a stent is to provide mechanical support in blood vessels, the mechanical properties of as-printed stents were evaluated comprehensively. First, the compressive and tensile behaviors were investigated by introducing linear deformation to the stent along its radial direction (inset of Figure 3a and b). In the compressive test, the stress-strain curve showed an almost linear relationship at the first ~30% strain (Figure 3a). The slop within this range yielded a compressive modulus of 37.7 kPa. The stent exhibited a quicker yield as the strain exceeded 35% and reached a failure point at 70% strain. The tensile performance was characterized by introducing a pair of internal bars inside the stent to introduce uniform tensile stress around the entire length of stent (inset of Figure 3b). Almost the entire stress-strain curve showed a good linearity with a slop of 32.0 kPa (from 5–75% strain) before failure, corresponding to the stent’s tensile modulus (Figure 3b). The most linear strain-stress relationships and the large failure tolerance (>70%) under both compressive and tensile conditions suggested that the as-printed stents are suitable for regular stents applications.²⁷

Figure 3.

Mechanical property of as-printed stents. (a, b) Strain-stress curves measured from as-printed stents under compressive (a) and tensile (b) conditions. Insets are photos of the corresponding mechanical testing setups. (c, d) Uniaxial cyclic compressive (c) and tensile (d) tests at various strain rate from 5% to 30% per minute. (e) Compressive loading-unloading behavior at a strain rate of 20% per minute for 20 cycles. (f) Long-term cyclic compressive loading-unloading behavior at a strain rate of 20% per minute for 100k cycles.

The strain rate influences on the stents’ mechanical property were then evaluated from uniaxial cyclic compressive and tensile tests at a series of strain rates ranging from 5% to 30% per minute (Figure 3c, d). A typical hysteresis could be observed from each compressive strain cycle (Figure 3c), revealing the viscoelastic behavior of the stents. The hysteresis loops slightly moved to the larger strain side as the strain rate and amplitude increased, suggesting a possible strain softening phenomenon²⁸ of the structure at large displacements. At a lower strain range from 5%−20% per minute, the stent showed a similar modulus of 39.9 kPa at different strain rates. Similar strain-related cycling behavior was observed from the series of tensile strains from 5% to 30% per minute (Figure 3d). Within the entire range of tensile strain rate from 5 to 30% per minute, all hysteresis loops exhibited a nearly identical starting point with very close slopes, suggesting the stable mechanical behavior under tensile deformations. At last, the cycling loading-unloading behavior of the stents were tested to evaluate its long-term performance under constant compressive straining actions, representing the practical stent application situations. Considering the intimal average circumferential compressive strain of stent was up to 13.1%²⁷ and that of physiological blood vessels under 100 mmHg was 5.9%,²⁹ a maximum 20% compressive strain was chosen in the cycling tests at a constant strain rate of 20% per minute. The cycling loading-unloading strain-stress curves showed a perfect match from continuous 20 strain cycles (Figure 3e), suggesting a good repeatability of the mechanical performance of stent when compressed repeatedly to its maximum strain range. The long-term cyclic loading-unloading mechanical stability of the stent was evaluated by performing the same cycling test over a much longer period. The strain-stress cyclic curves were measured at different cycling numbers from 0 to 100K after one-day relaxation (Figure 3f). All the curves showed a similar shape, suggesting the stent was able to retain its desired mechanical flexibility and stability over a long operation period. These analyses proved the potential of applying the 3D printed stents in the position of a regular stent to provide mechanical support to blood vessels.

The biocompatibility of the stent material (KNN/PVDF-HFP composite) is an essential property for safe in vivo applications. The biocompatibility was first examined on as-printed KNN/PVDF-HFP films using mouse aortic smooth muscle (MOVAS) cells. Immunofluorescence staining images revealed that cells incubated together with KNN/PVDF-HFP for 24, 48 and 72 hours exhibited the expected elongated morphology, along with appropriate distributions and densities. Importantly, there were no discernible differences compared to the cells in the control group, which was incubated under the same condition but without the presence of KNN/PVDF-HFP (Figure 4a). Additionally, a CellTiter-Glo (CTG) assay was conducted to quantitatively measure the viability of MOVAS cells in contact with the stent material. The results revealed that over 3 days of incubation, the average cell viability with KNN/PVDF-HFP was ~104.7%, showing no significant deviation from the control groups (100% viability) (Figure 4b). These findings affirmed that the KNN/PVDF-HFP stent material is nontoxic and biocompatible.

Figure 4.

In vitro biocompatibility test of the stent material. (a) Confocal imaging of MOVAS cells after 24h, 48h and 72h from the control and KNN/PVDF-HFP experimental groups, and (b) Cell viability measured by the CellTiter Glo assay from the control, and KNN/PVDF-HFP exposed group after 24h, 48h and 72h. The relative cell viability was expressed as (luminescence of sample wells - blank) / (luminescence of control wells - blank) × 100%, where blank is the luminescence of PBS plus CellTiter-Glo solution.

The piezoelectric output from the stents was eventually characterized under controlled strains in simulated pressurized conditions. The measurement design is schematically illustrated in Figure 5a. In the setup, a stent was sheathed on a flexible and gold coated PDMS tube (Figure S6). The gold surface (connecting the inner surface of the stent) was covered by an additional layer of insulating PDMS, and another gold layer was coated on the outer surface of the stent (inset of Figure 5a). Electrical output was measured from these two gold layers, representing the piezoelectric potential between the inner and outer surfaces of the stent. The stent-loaded and packaged PDMS tube was connected to a syringe pump integrated with a computer-controlled actuator, which was utilized to control the internal pressure at a designed rate (the system is shown in Figure S7a, Supporting video S3). The internal pressure change could drive the expansion and recovery of the PDMS tube, which could induce subtle strains to the stent simulating the movements in response to blood pressure fluctuations. Under a constant fluid pressure of 20 Psi, the stent produced a stable voltage output with a peak-to-peak value of ~0.07 V (Figure 5b). A zoom-in image of a single pair of voltage peaks that were separated by ~0.7 s, corresponding to the expansion and recovery motion of the PDMS tube (Figure 5c). Positive and negative voltage outputs can be detected respectively during pump-in fluid (vasodilation) and pump-out fluid (vasoconstriction) process. The slight peak intensity difference could be attributed to the thick PDMS tube wall, which might exhibit slower rate for shape recovery as the pressure reduced to the baseline. The peak polarization reversed as the connections to the inner and outer stent surface switched (Figure S8), confirming the output was a result of the directional piezoelectric effect. The electrical outputs also showed a good accordance with the pressure difference (ΔP) and increased monotonically as the pressure difference increased. At each ΔP value, the voltage output remained a stable pattern indicating a reliable piezoelectric performance of the stent (Figure 5d). The average peak-to-peak voltage increased from 0.07 V to 0.15 V as ΔP increased from 20 to 120 Psi, suggesting the potential to sensing blood pressure change. Long-term stability of the piezoelectric performance was further evaluated by monitoring the electrical outputs while the stent was strained repeatedly over 10,000 cycles. As shown in Figure 5e, the electrical outputs were measured every 2,500 cycles and maintained within a small range of 0.21–0.25 V, revealing a stable piezoelectric function of the piezoelectric stent over a long operation period.

Figure 5.

Piezoelectric output of the piezoelectric stents. (a) Schematic illustration of blood vessel straining simulation system for piezoelectric output measurement. Inset is a gold electrode-coated stent for the measurement. (b) Voltage output of piezoelectric stents under a pressure change (ΔP) from 0 of 20 Psi. (c) Zoom-in curve of a single pair of voltage peak from (b). (d) Voltage outputs of the piezoelectric stent under different ΔP. The average peak-to-peak voltage is also plotted as a function of ΔP showing a good linear relationship with R2=99.68%. (e) Long-time piezoelectric output test of a stent driven by ~10000 cycles at a constant ΔP of 50 Psi.

4. Conclusion

In summary, an intricate zigzag shaped piezoelectric stent was fabricated by FDM 3D printing assisted by in situ poling and rotation. The stent was made from a composite of ferroelectric KNN microparticles in a PVDF-HFP polymer matrix. Owing to the good homogeneity of the KNN/PVDF-HFP mixture, the as-printed stents exhibited desired mechanical property, biocompatibility, and piezoelectricity. The zigzag morphology offered a compressive modulus of ~37.7 kPa and a tensile modulus of ~32.0 kPa, demonstrating a capability of being compressed and expanded for potential catheter deliver. Long-term cyclic compression-release test revealed a good stability over 100 thousand straining cycles. Cell viability tests on vascular smooth muscle cells verified that there was no significant difference between control group and KNN/PVDF-HFP stent material, suggesting a good biocompatibility. Using a simulated pressurized system, the piezoelectric voltage outputs were measured directly from the inner and outer surfaces of the stent in response to internal pressure changes. The voltage output showed good accordance to pressure variations and sustained a stable output over long-term straining actions. This development demonstrated a successful strategy of introducing an additional electromechanical coupling function to stents, suggesting an intriguing future research direction for in vivo blood pressure monitoring or electricity generation solely from local blood pressure variations in blood vessels. This function may open a venue toward self-powered electrical stimulations to blood vessels from implanted stents, for wound healing or thrombosis prevention.