A method for preparing highly flexible and uniform piezoelectric nanoparticle assemblies and their application in force sensing.

Abstract

This invention discloses a method for preparing a highly flexible and uniform piezoelectric nanoparticle assembly and its application in force sensing, belonging to the field of sensing technology. The piezoelectric nanoparticle assembly achieves directional particle deposition through solvent evaporation and the adhesive elasticity of the substrate, forming a densely stacked grid morphology in PDMS. After polarization in an 80℃ oil bath, the ferroelectric domains are uniformly aligned, and the PDMS molecular chains encapsulate the barium titanate particles through van der Waals forces, effectively suppressing aggregation and cracking, ensuring the stability of the electromechanical coupling. The prepared grid piezoelectric structure can still output a stable peak voltage of 7.3 V after 10,000 bending cycles, with a voltage coefficient of 549 × 10⁻⁶. ‑3 V·m·N ‑1 This structure has significant application prospects in fields such as biomedical monitoring, wearable electronic skin, and soft robot sensing.

Description

Technical Field

[0001] This invention belongs to the field of sensing technology, specifically relating to a method for preparing a highly flexible and uniform piezoelectric nanoparticle assembly and its application in force sensing. Background Technology

[0002] The information disclosed in this background section is intended only to enhance understanding of the overall background of the invention and is not necessarily to be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art.

[0003] Multidimensional force sensing technology captures force and torque vector information in three-dimensional space, serving as the core tactile feedback system for robots. This technology facilitates the shift from position control to force / position hybrid control, effectively addressing damage and error issues caused by physical interference in contact-based operations. By decoupling and feeding back the interactive loads of the end effector in real time, multidimensional force sensing technology enables devices to possess active compliance—a characteristic that not only ensures operational accuracy but also enhances the safety of human-robot collaboration.

[0004] Piezoelectric structures, with their inherent electromechanical conversion characteristics and anisotropic response, naturally possess multidimensional force sensing capabilities. Compared to traditional piezoresistive or capacitive sensors, these structures can capture fine mechanical features under high-frequency vibration and transient impacts, and exhibit superior frequency response bandwidth. By utilizing the spatial coupling effect of the piezoelectric coefficient or specific topological designs, piezoelectric structures can decompose complex spatial vector forces into orthogonal axial and tangential components. This multidimensional force sensing technology not only extracts multiaxial loads with high accuracy but also provides a high signal-to-noise ratio and self-powered passive characteristics during complex dynamic interactions.

[0005] In recent years, with the continuous development of nanoparticle assembly technology, novel substrate materials, and controllable growth technology, researchers have employed various directional assembly processes to control the spatial arrangement of inorganic piezoelectric particles, enabling them to simultaneously respond to dynamic mechanical signals of different directions, types, and amplitudes, thereby meeting the complex application requirements of multimodal dynamic force sensing. Existing technologies include the directional assembly of lead zirconium phosphate (PZT) nanoparticles via dip-coating, producing a PZT structure with its c-axis parallel to the substrate. Other existing technologies disclose the construction of a multimodal dynamic force sensing sensor using directionally arranged barium titanate nanoparticles, employing chemical etching to increase particle content and optimize the interfacial bonding strength between the particles and the polymer matrix. Furthermore, existing technologies combine hydrothermal methods with electrodeposition techniques, precisely controlling reaction temperature and current density to achieve the growth of zinc oxide nanoparticles perpendicular to the substrate, producing large-area, highly uniform zinc oxide nanosheet arrays. However, inorganic piezoelectric particles exhibit poor flexibility and are prone to detachment from the substrate, making stable application in the field of dynamic force sensing difficult. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the purpose of this invention is to provide an application of piezoelectric nanoparticle assemblies in multidimensional and multimodal force sensing.

[0007] To achieve the above objectives, the technical solution of the present invention is as follows: In a first aspect, the present invention provides a method for preparing a highly flexible and uniform piezoelectric nanoparticle assembly, comprising the following steps: The solution-direct writing printing technology is used to directly write barium titanate onto an uncured elastic substrate. After printing, it is cured by heating and polarization to obtain the final product.

[0008] Secondly, the present invention provides a highly flexible and uniform piezoelectric nanoparticle assembly, which is prepared by the above-described preparation method.

[0009] Thirdly, the present invention provides a multi-dimensional force sensing device, comprising the above-mentioned highly flexible and consistent piezoelectric nanoparticle assembly.

[0010] Fourthly, the present invention provides the application of the above-mentioned highly flexible and consistent piezoelectric nanoparticle assembly or the above-mentioned multi-dimensional force sensing device in multi-dimensional and multi-modal force sensing.

[0011] Fifthly, the present invention provides a method for multidimensional force sensing, which employs the above-mentioned highly flexible and consistent piezoelectric nanoparticle assembly or the above-mentioned multidimensional force sensing device.

[0012] One or more of the above technical solutions have the following advantages or beneficial effects: This invention combines solution printing technology to directly print barium titanate ink onto a liquid PDMS substrate, thereby constructing a piezoelectric nanoparticle assembly (i.e., a flexible piezoelectric structure). This structure achieves directional particle deposition through solvent evaporation and the adhesive elasticity of the substrate, forming a densely stacked grid morphology within the PDMS. After polarization in an 80°C oil bath, the ferroelectric domains are uniformly aligned, and the PDMS molecular chains encapsulate the barium titanate particles through van der Waals forces, effectively suppressing aggregation and cracking, ensuring the stability of the electromechanical coupling. The prepared grid piezoelectric structure still outputs a stable peak voltage of 7.3 V after 10,000 bending cycles, with a voltage coefficient of 549 × 10⁻⁶. -3 V·m·N -1 This structure can accurately detect physiological signals such as joint movements, making it suitable for tactile perception in soft robots. It also features stress amplification and mechanical energy harvesting capabilities, and has significant application prospects in wearable electronic skin and soft robot sensing.

[0013] This invention embeds ink into PDMS to form an encapsulation structure through dynamic liquid-liquid interface assembly technology. This structure not only improves flexibility but also effectively prevents particle shedding, providing a feasible solution for multimodal force sensing devices. Attached Figure Description

[0014] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0015] Figure 1 This document presents a flowchart of the fabrication process for flexible piezoelectric structures and their related applications. Specifically, a) is a schematic diagram of direct-write printing technology and the process flow for fabricating flexible piezoelectric structures for application in the field of intelligent sensing; b) shows the application of this piezoelectric structure in the field of spine research; c) shows the application of this piezoelectric structure in the field of morphology sensing; and d) shows the application of this piezoelectric structure in the field of multi-dimensional vibration sensing of airflow / wind energy. Figure 2 The images show samples of barium titanate ink and related test results; where a represents barium titanate ink of different concentrations (30 wt%, 40 wt%, and 50 wt%), b represents viscosity curves of barium titanate ink of different concentrations, c represents Raman spectra of barium titanate ink of different concentrations, and d represents scanning electron microscope (SEM) images of barium titanate ink of different concentrations. Figure 3The diagrams show schematics of barium titanate ink embedded in the structure and their corresponding SEM images; where a is a schematic diagram of barium titanate ink embedded in the structure, b is a schematic diagram of barium titanate ink partially embedded in the structure, c is a schematic diagram of barium titanate ink not embedded in the structure, d is a schematic diagram of barium titanate ink fully embedded in the structure, e is a schematic diagram of barium titanate ink partially embedded in the structure, f is a schematic diagram of barium titanate ink not embedded in the structure, g is a cross-sectional microscope image of a fully embedded barium titanate ink structure, h is a cross-sectional microscope image of a partially embedded barium titanate ink structure, and i is a cross-sectional microscope image of a barium titanate ink structure not embedded in the structure. Figure 4 Optical microscope images of ink lines under different printing parameters: a) optical microscope image of ink lines at 7 psi pressure, 60 mm / s speed, and 0.5 mm height; b) optical microscope image of ink lines at 7 psi pressure, 60 mm / s speed, and 1 mm height; c) optical microscope image of ink lines at 7 psi pressure, 60 mm / s speed, and 1.5 mm height; d) optical microscope image of ink lines at 7 psi pressure and 30 mm / s speed; e) optical microscope image of ink lines at 8 psi pressure and 30 mm / s speed; f) optical microscope image of ink lines at 9 psi pressure and 30 mm / s speed; g) optical microscope image of ink lines at 7 psi pressure and 40 mm / s speed; h) optical microscope image of ink lines at 7 psi pressure and 50 mm / s speed; i) optical microscope image of ink lines at 7 psi pressure and 60 mm / s speed. Figure 5 The output voltage cycle stability test graphs of the piezoelectric structures with different embedding structures of barium titanate ink are shown. Among them, a is the output voltage graph of the fully embedded BaTiO3 structure after 1,000,000 cycles, b is the output voltage graph of the partially embedded BaTiO3 structure after 100,000 cycles, and c is the output voltage graph of the unembedded (surface attached) BaTiO3 structure after 100,000 cycles. Figure 6 The figures show the piezoelectric structure, XRD spectrum, and voltage stability before and after polarization; where a is the piezoelectric structure before polarization; b is the piezoelectric structure after polarization; c is the XRD spectrum before polarization; d is the XRD spectrum after polarization; e is the voltage stability of different embedded piezoelectric structures before polarization under tensile and bending conditions; and f is the voltage stability of different embedded piezoelectric structures after polarization under tensile and bending conditions. Figure 7The images show physical diagrams of three flexible piezoelectric structures and their corresponding long-term cyclic output stability test results. Specifically, a is a physical diagram of a single linear flexible piezoelectric structure, b is a physical diagram of a one-dimensional parallel array composed of five linear flexible piezoelectric structures, and c is a physical diagram of a two-dimensional grid array composed of intersecting linear flexible piezoelectric structures. d, e, and f correspond to the output voltage variation curves of the three structures (single unit, one-dimensional array, and two-dimensional grid array) under 1,000,000 mechanical cycles, respectively. Figure 8 A schematic diagram illustrating the polarization process and piezoelectric working mechanism of a flexible piezoelectric structure; Figure 9 The figures represent the dynamic response performance of five single-wire flexible devices under different pressure excitations. Among them, ae corresponds to the physical images of these five single-wire flexible devices. The corresponding I shows the curve of the output voltage changing with time when the device is subjected to a force of 0.1N and II shows the curve of the output voltage changing with time when the device is subjected to a force of 1N. Figure 10 The figures represent the dynamic response performance of five single-wire flexible piezoelectric devices under different pressure excitations; where ae correspond to the physical images of these five single-wire flexible devices respectively; I shows the curve of the output voltage changing with time when the device is subjected to a force of 0.1N and II shows the curve of the output voltage changing with time when the device is subjected to a force of 1N. Figure 11 Schematic diagrams of linear or grid-like flexible piezoelectric structures and corresponding flexible piezoelectric structures bent along or perpendicular to the printing direction, and the voltages they generate; wherein, a is a schematic diagram of a linear flexible piezoelectric structure, b is a schematic diagram of a linear flexible piezoelectric structure bent along the printing direction and the voltages it generates, c is a schematic diagram of a linear flexible piezoelectric structure bent perpendicular to the printing direction and the voltages it generates, d is a schematic diagram of a grid-like flexible piezoelectric structure, e is a schematic diagram of a grid-like flexible piezoelectric structure bent along the printing direction and the voltages it generates, and f is a schematic diagram of a grid-like flexible piezoelectric structure bent perpendicular to the printing direction and the voltages it generates. Figure 12 The diagrams show physical images of flexible devices with different structures and their corresponding electrical performance test diagrams. Among them, a is a physical image of a single linear flexible piezoelectric structure, b is a physical image of a one-dimensional parallel array flexible device, and c is a physical image of a two-dimensional cross-grid array flexible device. I is the dynamic voltage response diagram of the corresponding flexible device under a step force of 0-01N, II is the pressure-voltage relationship diagram of the corresponding flexible device with a structure of 0-1kPa, and III is the response time diagram of the corresponding flexible device with a structure of 1N under a force. Figure 13The images show physical diagrams of flexible devices with different structures and their corresponding electrical performance test diagrams; where a is a physical diagram of a single linear flexible piezoelectric structure, b is a physical diagram of a one-dimensional parallel array flexible device; I is the pressure-voltage relationship diagram of the corresponding flexible device with 0-1 kPa, and II is the voltage-time response diagram of the corresponding flexible device with 0.1 N force. Figure 14 The results of using flexible piezoelectric structures for spinal detection are shown; where a represents five different postures of the spine during movement; bd represents the piezoelectric signals output by the piezoelectric structures attached to the upper, middle and lower parts of the spine, respectively, during spinal movement monitoring. Figure 15 The experiment identifies the spine and its voltage changes when a flexible piezoelectric structure is attached to a chair, and generates scatter plots and confusion matrices. Specifically, a) is a schematic diagram of the flexible piezoelectric structure attached to the chair back; b) shows the voltage generated by the flexible piezoelectric structure when the human body is in an upright posture; c) shows the voltage generated by the flexible piezoelectric structure when the human body is in a relaxed posture; d) shows the voltage generated by the flexible piezoelectric structure when the human body is in a slumped posture; e) is an image of the sensor attached to the human throat; f) shows the voltage response when the test subject pronounces "a", "e", and "i"; g) shows the voltage response when the test subject pronounces "l", "m", and "n"; h) is a t-SNE feature clustering scatter plot; and i) is the confusion matrix of the flexible piezoelectric structure signal. Figure 16 The image shows a simulation of the human body using a flexible piezoelectric structure to identify the spine and its voltage changes. From left to right, the images show the human body in an upright posture, a relaxed posture, and a slumped posture. Figure 17 This paper presents the application of flexible piezoelectric structures in the detection of irregular surface morphology based on machine learning. In this diagram, a shows a flexible piezoelectric structure under pressure on a flat surface and the resulting voltage; b shows a flexible piezoelectric structure under pressure on a non-flat surface and the resulting voltage. Figure 18 The diagram shows the voltage generated by a flexible piezoelectric structure as a tactile sensor in both planar and non-planar environments. Specifically, a represents the voltage generated by the flexible piezoelectric structure as a tactile sensor in both planar and non-planar environments; b represents the voltage generated by the flexible piezoelectric structure as a tactile sensor in both planar and non-planar environments. Figure 19The diagrams show the effects of a flexible piezoelectric structure attaching to and sliding across a paper-based plane or a non-flat surface as a form of tactile sensing, voltage changes, confusion matrices, and scatter plots. Specifically, a represents the effect of a flexible piezoelectric structure attached to a paper-based plane as a form of tactile sensing and voltage changes; b represents the effect of a flexible piezoelectric structure sliding across a paper-based plane as a form of tactile sensing and voltage changes; c represents the effect of a flexible piezoelectric structure sliding across a non-flat surface as a form of tactile sensing and voltage changes (I, II, III, IV, and V represent the pressure signals generated by each tentacle); d represents the confusion matrix of the flexible piezoelectric structure signals; and e represents the t-SNE feature clustering scatter plot. Figure 20 This is an application of flexible piezoelectric structures in vibration sensing; where a is the sensor's response to sound stimuli at different decibel levels (30~60 dB), and its voltage output signal changes with time and sound intensity; b is the characteristic voltage waveform generated by the door opening action; c is the characteristic voltage waveform generated by the window opening action; d is the t-SNE visualization map of the sensor signal characteristics; and e is the confusion matrix of the binary action classification task. Figure 21 A schematic diagram of ten branches of a flexible piezoelectric structure; Figure 22 The diagram shows the force and voltage variations of a flexible piezoelectric structure under different branches, along with scatter plots and confusion matrices. Specifically, a represents the force and voltage variations of the ten branches of the flexible piezoelectric structure under tilting directions; b represents the force and voltage variations of the five branches under positive directions; c represents the force and voltage variations of the five branches under bottom directions; d is a scatter plot of t-SNE feature clustering; and e is the confusion matrix of the flexible piezoelectric structure signal. Figure 23 Applications of intelligent display technology and lighting devices; where, a) applications of intelligent display technology: (I) a scheme to light up different numbers of bulbs by pressing with a finger; (II) different numbers of bulbs with flexible piezoelectric structures as intelligent displays; (III) bulbs with different digital information that are lit by pressing with a finger as intelligent displays; b) applications of lighting devices: (I) handheld lighting devices with flexible piezoelectric structures; (II) handheld ambient lighting devices with flexible piezoelectric structures; (III) wearable lighting devices with flexible piezoelectric structures; Figure 24The diagram shows the assembled structure, the controllable output voltage under different numbers of printed circuit lines, and the tip discharge phenomenon. Specifically, a) is a schematic diagram of a flexible piezoelectric structure that achieves controllable high-voltage output of tip discharge using machine learning; b) shows the controllable output voltage of a flexible piezoelectric structure with a length of 4 cm and a diameter of 40 μm under different numbers of printed circuit lines: (I) measured results under different surface pressures; (II) prediction results based on MLP neural networks under surface pressure; (III) prediction results based on MLP neural networks under a force of 500 N; c) shows the tip discharge phenomenon of the flexible piezoelectric structure: (I) tip discharge of multiple flexible piezoelectric structures connected in series in humid air; (II) multiple flexible piezoelectric structures connected in series in ethanol gas. Figure 25 It is a barium titanate ink with a concentration of 40 wt%, suitable for direct writing and printing; Figure 26 To directly write the actual image of the printed piezoelectric structure; Figure 27 A schematic diagram of a piezoelectric structure undergoing cyclic bending; Figure 28 The diagram shows a flexible piezoelectric structure and its overall / separate sensing functions as a tentacle; where a is a schematic diagram of the flexible piezoelectric structure as a separate sensing function of the tentacle, b is a schematic diagram of the flexible piezoelectric structure as a whole sensing function of the tentacle, and c is a schematic diagram of the flexible piezoelectric structure. Figure 29 The diagram illustrates the use of a flexible piezoelectric structure attached to a paper / metal substrate as a tactile sensor and the corresponding voltage changes. Specifically, a represents a flexible piezoelectric structure attached to a paper substrate as a tactile sensor, b represents the overall voltage change of the flexible piezoelectric structure attached to a paper substrate as a tactile sensor, c represents a flexible piezoelectric structure attached to a metal substrate as a tactile sensor, and d represents the overall voltage change of the flexible piezoelectric structure attached to a metal substrate as a tactile sensor. Figure 30 The diagram shows the voltage generated by the flexible piezoelectric structure as a tactile sensor in the same / different planes. Among them, a is the voltage generated by the flexible piezoelectric structure as a tactile sensor in the same plane. b is the voltage generated by the flexible piezoelectric structure as a tactile sensor in different planes. Figure 31 The diagrams show the pressure and voltage generated on flexible piezoelectric structures under the same / different planes; where a is the pressure and voltage generated on the flexible piezoelectric structure under the same plane, and b is the pressure and voltage generated on the flexible piezoelectric structure under different planes. Detailed Implementation

[0016] This invention employs direct-write printing technology to fabricate piezoelectric nanoparticle assemblies for multidimensional force sensing. Barium titanate nanoparticle ink is directly written onto a liquid, uncured polydimethylsiloxane (PDMS) substrate. As the ink solvent evaporates, the barium titanate nanoparticles undergo a dynamic assembly process due to the shrinkage of the uncured PDMS substrate. Subsequently, the compact barium titanate nanoparticle assemblies are securely encapsulated within the cured PDMS. The prepared BaTiO3 nanoparticle flexible piezoelectric structure is polarized by applying a polarization voltage of 450V in an 80°C oil bath. The encapsulated barium titanate nanoparticle assemblies exhibit a consistent piezoelectric response, with a piezoelectric coefficient ranging from 37.3 pC·N. -1 Increased to 51.2 pC·N -1 At 2.8 cm -1 After undergoing 10,000 bending cycles at the curvature, the film still outputs a stable piezoelectric signal. Through structural separation, the minimum detectable force of the film is increased from 0.1 N for the entire film to 0.01 N for a single linear unit, and the sensitivity is also increased from 0.63 V·N for the entire film. -1 Increased to 5.6 V·N per linear unit. -1 Leveraging the advantages of direct-write printing technology in controlled patterning, large-scale processing, and independent uniform preparation, this encapsulated barium titanate nanoparticle assembly can be used for multi-dimensional force sensing applications such as analyzing human motion, detecting irregular surface morphologies, and recognizing sound signals using machine learning.

[0017] This invention aims to develop novel piezoelectric structures and leverages the precise molding advantages of solution-to-print technology to successfully fabricate flexible piezoelectric structures on uncured PDMS substrates by directly writing and printing barium titanate. Barium titanate ink achieves directional particle deposition through the synergistic effect of solvent evaporation and substrate viscoelasticity matching, ultimately forming a densely stacked mesh-like flexible piezoelectric structure on the cured PDMS substrate. After oil bath heating polarization treatment at 80℃ and 450V polarization voltage, the ferroelectric domains of the barium titanate nanoparticles align in the same direction. The oil bath environment reduces the rotational resistance between domains, facilitating their regular alignment under the influence of an external electric field. PDMS molecular chains are adsorbed onto the barium titanate surface through van der Waals forces, effectively suppressing particle aggregation and structural cracking, and maintaining a uniform distribution of the piezoelectric phase during electromechanical coupling. A flexible piezoelectric structure with an 11×11 grid was printed on a 4.4×4.4 cm cured PDMS substrate with a 4×4 mm spacing. After 10,000 bending cycles, it still stably outputs a peak voltage of 7.3V, demonstrating excellent structural stability and piezoelectric durability. This directly write-printed flexible piezoelectric structure combines the functions of a sensing element and a medical component, and can be widely used in real-time monitoring of spinal motion, force sensing monitoring of irregular shapes, and intelligent recognition monitoring of different letters, providing a practical and feasible technical path for the development of flexible electronics and force sensing fields.

[0018] The piezoelectric nanoparticle assemblies exhibit uniformity. The piezoelectric nanoparticles in the ink possess uniformity. The uniformity of threshold-controlled programmed assembly endows the structure with an embedded morphology. Polarization modulates the uniformity of piezoelectric properties.

[0019] The piezoelectric nanoparticle assembly exhibits high flexibility and uniformity. Regarding flexibility, the structure utilizes dynamic liquid-liquid interface assembly technology to embed barium titanate ink into PDMS to form an encapsulated structure. During mechanical deformation, the PDMS elastomer mesh can absorb mechanical energy and provide mechanical buffering through the flexible stretching of molecular chains. The piezoelectric nanoparticle assembly maintains structural stability even after repeated bending cycles. Regarding uniformity, firstly, the uniformity of the piezoelectric nanoparticles in the ink forms the underlying foundation. The barium titanate nanoparticles prepared by the hydrothermal method are uniform in size and have high crystallinity, achieving high dispersion and stable suspension in the ink system, thus preventing particle aggregation and phase separation at the source and ensuring absolute uniformity of the functional phase in spatial distribution. Secondly, the uniformity of threshold-limited programmed assembly endows the structure with its embedded morphology. By precisely controlling the rheological threshold properties and spatial printing trajectory of the ink within the liquid PDMS substrate, highly repeatable and precise control over the line width, interlayer spacing, and overall morphology of the embedded array is achieved, perfectly eliminating deformation and defects during structural forming. Ultimately, the dual uniformity of material distribution and spatial structure is effectively transmitted to the consistency of piezoelectric performance regulated by polarization. When a polarization electric field is applied, the uniform distribution of the local electric field within the structure is ensured, prompting the barium titanate ferroelectric domains to achieve synchronous and highly consistent directional flipping. This endows the device with stable, reliable, and highly consistent piezoelectric response and electromechanical coupling performance in different macroscopic regions, different stress points, and long-cycle periods.

[0020] In one typical embodiment, the present invention provides a method for preparing a highly flexible and uniform piezoelectric nanoparticle assembly, comprising the following steps: The solution-direct writing printing technology is used to directly write barium titanate onto an uncured elastic substrate. After printing, it is cured by heating and polarization to obtain the final product.

[0021] In some embodiments of this implementation, during printing, the pattern design employs a single element, a one-dimensional array, or a two-dimensional grid array; the one-dimensional array is a piezoelectric array composed of multiple single-line structures (including one-dimensional parallel arrays), which includes radially arranged parallel-head tentacles (such as...). Figure 22 On the left side, the branches are aligned at one end, and at the other end, multiple branches radiate outwards in the form of tentacles) or in a ring-shaped branching pattern (such as...). Figure 22 On the right side, multiple branches are connected together in a ring at one end, and radiate outwards in the form of tentacles at the other end. The number of tentacles / branches is not specifically limited and can be adjusted according to the actual application, such as 1 to 20, preferably 5 to 10.

[0022] In some embodiments of this implementation, the preparation method of the piezoelectric nanoparticle assembly includes the following steps: using solution direct writing printing technology, a flexible piezoelectric structure is prepared on an uncured PDMS substrate by direct writing and printing barium titanate, and a polarization voltage of 450V is applied in an 80°C oil bath environment to polarize the prepared piezoelectric nanoparticle assembly.

[0023] Specifically: A liquid viscoelastic substrate is fixed on a printing station, barium titanate ink is injected into the cartridge, and direct writing printing is performed. The printed sample is then heated and cured to obtain a piezoelectric nanoparticle assembly. Further, the specific process of direct writing printing is as follows: The cartridge uses a 100-200 micrometer (preferably 150 micrometer) printing needle, and direct writing printing is performed according to a preset program at a pressure of 5-20 psi (specifically 5psi, 6psi, 7psi, 8psi, 9psi, 10psi, 11psi, 12psi, 15psi, 18psi, 20psi, etc., preferably 7-10 psi). The printing speed is 30-60 mm / s, specifically 30 mm / s, 40 mm / s, 50 mm / s, 60 mm / s, etc., and the height is 0-1.5 mm, specifically 0 mm, 0.5 mm, 1 mm, 1.5 mm, etc. Furthermore, the heat curing conditions are: heat curing at 80~90℃ for 30~50 minutes, preferably heat curing at 85℃ for 40 minutes.

[0024] In some embodiments of this implementation, the preparation of barium titanate nanoparticle ink involves using barium titanate nanoparticles as a functional phase precursor and a homogeneous mixture of dimethyl sulfoxide and ethylene glycol as a composite dispersion solvent. The two are mixed to obtain a barium titanate precursor dispersion, which is then subjected to ultrasonic treatment to obtain the final product.

[0025] Furthermore, the preparation process of barium titanate nanoparticles is as follows: barium nitrate, oleic acid, oleylamine, Ti(OBu)4, and alkaline water are mixed and subjected to a hydrothermal reaction. The hydrothermal reaction conditions are a homogeneous reaction at 140~160℃ for 10~20 h.

[0026] Furthermore, the mass fraction of barium titanate nanoparticles is 30~50wt%.

[0027] Furthermore, in the composite dispersion solvent, the volume ratio of dimethyl sulfoxide to ethylene glycol is (0.9~1.1):(0.9~1.1).

[0028] Furthermore, the ultrasonic treatment conditions are: 0.5~5h at room temperature, preferably 0.5~1.5h, to promote the complete depolymerization and uniform dispersion of nanoparticles in the binary mixed solvent, thereby forming a stable colloidal system.

[0029] Preferably, the preparation of barium titanate nanoparticle ink includes the following steps: Dry barium titanate nanoparticles were used as the functional phase precursor, while a homogeneous mixture of dimethyl sulfoxide and ethylene glycol at a volume ratio of 1:1 was used as the composite dispersion solvent. In this study, formulations were designed based on mass fractions of 30 wt%, 40 wt%, and 50 wt%, and barium titanate nanoparticles were sequentially added to the aforementioned mixed solvent. After thorough stirring and preliminary mixing, barium titanate precursor dispersions with different solid contents were obtained. Subsequently, the prepared dispersions of each concentration were placed in an ultrasonic cleaning device and subjected to continuous ultrasonic treatment at room temperature for 1 hour to promote complete deagglomeration and uniform dispersion of the nanoparticles in the binary mixed solvent, thereby forming a stable colloidal system. The resulting functional barium titanate ink exhibits uniform dispersibility, good stability, and can be directly applied to direct-write printing processes.

[0030] In some embodiments of this implementation, the viscoelastic PDMS substrate is prepared by mixing a PDMS precursor with a curing agent to prepare liquid PDMS, and then spin-coating the liquid PDMS onto the surface of a pretreated PET substrate.

[0031] Preferably, the PDMS precursor and curing agent are mixed and reacted in a vacuum environment for 0.5 to 1 hour to remove air bubbles.

[0032] Preferably, the mass ratio of PDMS precursor to curing agent is (5~15):1, and more preferably (8~12):1.

[0033] Preferably, the PDMS precursor comprises vinyl-terminated polydimethylsiloxane. The curing agent is a polysiloxane containing silane-hydrogen bonds, most commonly polymethylhydrosiloxane.

[0034] Preferably, the pretreatment of the PET substrate includes: ultrasonic cleaning and drying in an alcohol container.

[0035] Preferably, the specific spin coating conditions are as follows: 280~320 rpm for 20~40 seconds, 380~420 rpm for 20~40 seconds, 480~520 rpm for 20~40 seconds, 580~620 rpm for 20~40 seconds, 680~720 rpm for 20~40 seconds, and 980~1020 rpm for 20~40 seconds. This stepped spin coating is a method for controlling the structure in the early stages. The most preferred conditions are: 300 rpm for 30 seconds, 400 rpm for 30 seconds, 500 rpm for 30 seconds, 600 rpm for 30 seconds, 700 rpm for 30 seconds, and 1000 rpm for 30 seconds.

[0036] The preparation of the viscoelastic PDMS substrate includes the following steps: In this study, PDMS precursor and curing agent were prepared at a mass ratio of 10:1. After thorough stirring, the mixture was reacted in a vacuum environment for 0.5 hours to remove air bubbles. PET boards were cut into two sizes: 4.4×4.4 cm and 1.2×1.2 cm, and placed in a container of ethanol for ultrasonic cleaning for 5 minutes. The cleaned PET boards were then placed in a vacuum drying oven and heated to 90°C until completely dry. Subsequently, the treated PET substrates were removed, and liquid PDMS was spin-coated onto their surface using a spin coater. The specific spin-coating conditions were as follows: 300 rpm for 30 seconds, 400 rpm for 30 seconds, 500 rpm for 30 seconds, 600 rpm for 30 seconds, 700 rpm for 30 seconds, and 1000 rpm for 30 seconds, ultimately yielding a liquid viscoelastic substrate.

[0037] In one typical embodiment, the present invention provides a highly flexible and uniform piezoelectric nanoparticle assembly, which is prepared by the above-described preparation method.

[0038] In one typical embodiment, the present invention provides a multi-dimensional force sensing device, comprising the above-described highly flexible and consistent piezoelectric nanoparticle assembly.

[0039] In one typical embodiment, the present invention provides the application of the above-described highly flexible and consistent piezoelectric nanoparticle assembly or the above-described multidimensional force sensing device in multidimensional and multimodal force sensing.

[0040] Independent or integrated sensing is employed. A modular tentacle array was constructed based on a flexible piezoelectric structure manufactured using direct-write printing technology. This array supports both independent and integrated sensing operating modes and provides a structural foundation for multidimensional tactile detection.

[0041] The piezoelectric nanoparticle assembly is an encapsulated piezoelectric nanoparticle assembly formed by encapsulating barium titanate nanoparticles in a viscoelastic PDMS substrate.

[0042] To achieve superior technical performance, a mesh-like flexible piezoelectric structure was employed, consisting of densely stacked mesh-like flexible piezoelectric structures formed on a cured PDMS substrate. This mesh structure uniformly disperses the stress generated during stretching, thus avoiding localized stress concentrations that can occur in linear structures. This uniform stress dispersion results in more uniform and stable lattice distortion of the barium titanate nanoparticles. Consequently, the electrical signal output by the mesh-like flexible piezoelectric structure remains stable during stretching, without significant fluctuations, further demonstrating the unique advantages of the mesh structure in stress regulation and piezoelectric performance stability.

[0043] After polarization is applied, the piezoelectric nanoparticle assembly exhibits a piezoelectric response. The polarization conditions are: oil bath heating polarization treatment at 70~90℃ and 400~500V, preferably 80℃ and 450V.

[0044] The multidimensional force sensing includes applications such as analyzing human movement, detecting irregular surface morphology, and recognizing sound signals using machine learning. The piezoelectric nanoparticle assembly structure can accurately sense continuous spinal movements. By sensing the dynamic stress applied to the spine and converting it into characteristic electrical signals, combined with t-SNE dimensionality reduction analysis and machine learning classification algorithms, it achieves highly sensitive capture and accurate intelligent recognition of various spinal movements. In tactile sensing mode, the piezoelectric nanoparticle assembly structure, acting as a "tentacle," demonstrates a comprehensive capability for both static accurate recognition and dynamic high-resolution tracking of complex surface morphologies. In static contact scenarios, the piezoelectric nanoparticle assembly structure can keenly capture and distinguish the differences in local multidimensional stress distribution between two-dimensional flat surfaces and three-dimensional stepped undulating surfaces by outputting a stable and consistent single-peak voltage or a significantly enhanced multi-peak voltage. During dynamic sliding interaction, when the piezoelectric nanoparticle assembly structure slides over paper-based or uneven surfaces, it can sensitively convert the continuous bending and rebound stress caused by alternating surface textures and continuous friction into multiple repetitive and highly consistent continuous multi-peak pulse signals. This piezoelectric nanoparticle assembly structure, acting as a tentacle, combined with t-SNE dimensionality reduction analysis and machine learning classification algorithms, achieved accurate identification of surface materials and textures. In vibration sensing scenarios, the piezoelectric nanoparticle assembly structure was used to distinguish between throat sound vibrations and acoustic sound vibrations. The ten-branched piezoelectric nanoparticle assembly structure can differentiate the local spatial multidimensional stress distribution differences caused by different vibrations. When the piezoelectric nanoparticle assembly structure is subjected to tilting, positive phase, and bottom-direction forces, wind forces cause different degrees of bending deformation in each branch, resulting in the output of electrical signals with significantly different characteristics. By combining t-SNE dimensionality reduction feature clustering analysis and machine learning classification algorithms, accurate and intelligent identification of environmental vibrations was successfully achieved.

[0045] With the development of intelligent technology, many devices require electrical energy to function. Therefore, the efficient harvesting of mechanical energy and its conversion into electrical energy can be applied to various intelligent fields, such as energy harvesting, self-powered systems, and intelligent displays. Piezoelectric structures possess the unique ability to efficiently harvest mechanical energy and convert it into electrical energy. Due to their non-centrosymmetric structure, piezoelectric structures exhibit electrode polarization under external forces. The amount of charge is proportional to the magnitude of the external force. Without relying on grid power or disposable batteries, piezoelectric structures can harvest mechanical energy such as pressure and torsion generated by human movement and convert it into electrical energy to power intelligent displays and lighting. Flexible piezoelectric structures can directly provide stable power support for lighting, warning, decorative, and display devices without grid power or disposable batteries. Leveraging their self-powered characteristics, flexible piezoelectric structures can be quantitatively controlled using machine learning methods to efficiently convert mechanical energy into controllable high voltage. Direct-write printing technology provides a solution for the patterned mass production of encapsulated piezoelectric structures, which can harvest mechanical energy from human movement for continuous and stable self-powered systems. Machine learning methods can quantify force-sensing performance, releasing controllable high voltages for chemical ionization detection.

[0046] In this invention, a piezoelectric nanoparticle assembly structure was prepared by printing BaTiO3 nanoparticle ink onto a liquid, uncured PDMS substrate. The self-powered performance of this assembly structure was investigated, revealing that due to its unique self-encapsulation characteristics and the presence of the piezoelectric effect, the structure can efficiently collect mechanical energy generated by human movement and convert it into electrical energy, exhibiting excellent energy conversion behavior. This makes it suitable for self-powered applications such as decorative lighting and smart displays. Simultaneously, a multilayer perceptron (MLP) neural network architecture was combined to achieve controllable high-voltage prediction and output. High-voltage discharge phenomena at the tip were observed by accurately predicting and cascading multiple piezoelectric nanoparticle assemblies, generating pale purple and pale blue sparks in humid air and ethanol gas, respectively, enabling the ionization, differentiation, and rapid screening of specific chemical gases. This piezoelectric nanoparticle assembly structure can efficiently collect mechanical energy generated by the environment and human movement, demonstrating significant application potential in self-powered sensing, wearable electronic devices, and smart displays. It provides a feasible approach for developing next-generation customized, highly durable, non-traditional self-powered power generation devices.

[0047] The multimodal force sensing includes self-powered sensors, wearable electronic devices, decorative lighting, and smart displays.

[0048] In one typical implementation, a multi-dimensional force sensing method employs the aforementioned highly flexible and consistent piezoelectric nanoparticle assembly or the aforementioned multi-dimensional force sensing device.

[0049] In this invention, unless otherwise specified, all other test materials and instruments are conventional test materials in the field and can be purchased through commercial channels.

[0050] Ethylene glycol (98%) was purchased from Shanghai Maclean's Biotechnology Co., Ltd. Dimethyl sulfoxide (99%) was purchased from Shanghai Maclean's Biotechnology Co., Ltd. Polyethylene terephthalate (PET, 200 μm thickness) was purchased from Beijing Daxiang Plastic Products Factory. Polydimethylsiloxane (PDMS, model DC184) was purchased from Dow Corning. Copper foil (0.1 mm thickness) was purchased from Strip GmbH, Germany. Ethanol (analytical grade AR, 95%) was purchased from Sinopharm Chemical Reagent Co., Ltd. Deionized water was purchased from Watson.

[0051] The preparation process of barium titanate nanoparticles includes: firstly, adding 10 mL of an aqueous solution containing 4 mmol Ba(NO3)2, 7.2 mL of oleic acid, 3.6 mL of oleylamine, and 40 mL of a n-butanol solution containing 4.8 mmol Ti(OBu)4 to 100 mL of Teflon-liner. Finally, slowly adding 10 mL of an aqueous solution containing 10 mmol NaOH, and then placing the Teflon-liner in a stainless steel autoclave. The reaction mixture was placed in a homogeneous reactor at 150 °C for 18 h. After cooling to room temperature, the product was collected using the 1-butanol phase and the aqueous phase to obtain a stable emulsion colloidal solution. After centrifugation at 4000 rpm to clarify the solution, it was heated in a vacuum drying oven to obtain BaTiO3 nanoparticles.

[0052] To enable those skilled in the art to better understand the technical solution of the present invention, the technical solution of the present invention will be described in detail below with reference to specific embodiments.

[0053] Example 1 1. Preparation process: (1) Preparation of barium titanate nanoparticle ink: Dry barium titanate nanoparticles were used as the functional phase precursor, while a homogeneous mixture of dimethyl sulfoxide and ethylene glycol at a volume ratio of 1:1 was used as the composite dispersion solvent. In this study, formulations were designed based on mass fractions of 30 wt%, 40 wt%, and 50 wt%, and barium titanate nanoparticles were sequentially added to the aforementioned mixed solvent. After thorough stirring and preliminary mixing, barium titanate precursor dispersions with different solid contents were obtained. Subsequently, the prepared dispersions of each concentration were placed in an ultrasonic cleaning device and subjected to continuous ultrasonic treatment at room temperature for 1 hour to promote complete deagglomeration and uniform dispersion of the nanoparticles in the binary mixed solvent, thereby forming a stable colloidal system. The resulting functional barium titanate ink exhibits uniform dispersibility, good stability, and can be directly applied to direct-write printing processes.

[0054] (2) Preparation of viscoelastic PDMS substrate: In this study, PDMS precursor and curing agent were prepared at a mass ratio of 10:1. After thorough stirring, the mixture was reacted in a vacuum environment for 0.5 hours to remove air bubbles. PET boards were cut into two sizes: 4.4×4.4 cm and 1.2×1.2 cm, and placed in a container of ethanol for ultrasonic cleaning for 5 minutes. The cleaned PET boards were then placed in a vacuum drying oven and heated to 90°C until completely dry. Subsequently, the treated PET substrates were removed, and liquid PDMS was spin-coated onto their surface using a spin coater. The specific spin-coating conditions were as follows: 300 rpm for 30 seconds, 400 rpm for 30 seconds, 500 rpm for 30 seconds, 600 rpm for 30 seconds, 700 rpm for 30 seconds, and 1000 rpm for 30 seconds, ultimately yielding a liquid viscoelastic substrate.

[0055] (3) Direct writing and printing technology for flexible piezoelectric structures: A liquid viscoelastic substrate was fixed on a printing station, and barium titanate ink was injected into an ink cartridge. The cartridge used a 150-micron printing needle, and direct writing printing was performed at a pressure of 10 psi according to a preset program. The printed sample was then heated and cured at 85°C for 40 minutes, and copper foil was attached to both ends. A polarization voltage of 450V was applied in an 80°C oil bath to polarize the prepared BaTiO3 nanoparticle flexible piezoelectric structure, promoting the uniform orientation of ferroelectric domains along the external electric field direction. This effectively improved the crystal order and spontaneous polarization intensity, ultimately forming a flexible piezoelectric structure.

[0056] (4) Characteristic analysis and measurement: The particle size of the nanoparticles was determined using a particle size analyzer (Mastersizer 3000), which provides quantitative data on particle distribution. Ink performance testing was performed using a rotational rheometer to characterize the ink's rheological properties. Nanoparticle images were acquired using an electron microscope (NP 800TRF) to reveal the morphological characteristics of the nanoparticles; a field emission scanning electron microscope (Ruguus 8220) was used to acquire SEM images, showcasing the microstructure of the samples. The piezoelectric crystal phase was identified using Raman spectroscopy to confirm the crystalline phase composition of the material. XRD diffraction patterns were measured using a Smart Lab SE X-ray diffractometer manufactured by Rigaku Corporation, employing Cu Kα radiation (λ = 0.15405 nm) with a scanning range of 5–80°. Bending tests were conducted on a custom bending stage from Anach, which applied repeated bending deformation to the samples; voltage changes were monitored in real-time using a digital source oscilloscope (Keithley 2400) connected to a computer, and the electrical response signal of the samples was recorded.

[0057] 2. Data Analysis: Physicochemical properties of barium titanate ink: This invention demonstrates inks containing 30 wt%, 40 wt%, and 50 wt% barium titanate, respectively, which correspond to different proportions of barium titanate nanoparticles in the dispersion system. Figure 2 (a) The content of barium titanate in the ink and its corresponding rheological properties are crucial for achieving accurate direct-write printing; suitable viscosity and dispersion state are the core prerequisites for ensuring stable ink extrusion and uniform printing structure. Barium titanate inks of different concentrations all exhibit typical shear-thinning rheological behavior: the ink viscosity gradually decreases with increasing shear rate, while the initial viscosity increases with increasing barium titanate content. The initial viscosity of the 50 wt% barium titanate ink is significantly higher than that of the 30 wt% sample, because the high particle content enhances the internal interaction resistance of the ink (…). Figure 2 (b) The crystalline phase of barium titanate nanoparticles constitutes the structural basis for their piezoelectric properties. Figure 2 (c) Raman spectroscopy results show that barium titanate inks of different concentrations all exhibit tetragonal phase characteristic absorption peaks (257 cm⁻¹). -1 306cm -1 514 cm -1 715 cm -1 This indicates that the nanoparticles retain a tetragonal phase structure with spontaneous polarization characteristics, providing a material basis for the performance of flexible piezoelectric structures. Electron microscopy characterization revealed the morphology and dispersion state of the barium titanate nanoparticles after ink dispersion: the quasi-spherical barium titanate nanoparticles exhibited a relatively uniform particle size distribution, maintaining good dispersibility and no obvious agglomeration at a moderate content of 40 wt%; however, when the barium titanate content increased to 50 wt%, some particle aggregation occurred. Figure 2 (d) In 50 wt% barium titanate ink, excessive nanoparticles will agglomerate under the influence of van der Waals forces, which will not only change the rheological properties of the ink and reduce the printing performance, but also interfere with the generation and orderly change of electric dipole moment in the subsequent piezoelectric structure, thus adversely affecting the piezoelectric performance.

[0058] SEM results showed that the BaTiO3 nanoparticles exhibited extremely high size uniformity and a uniform particle size distribution; Raman spectroscopy further verified the deep lattice state of the material structure. The spectrum clearly and sharply presented the typical tetragonal phase vibration peaks of BaTiO3. This significant crystal feature strongly confirms that during the initial ink preparation and physical stirring process, the internal lattice of BaTiO3 did not undergo a non-piezoelectric phase transformation or lattice destruction, and the piezoelectric structure remained intact.

[0059] Therefore, the viscosity, dispersion state, crystal phase, tetragonal phase structure with spontaneous polarization characteristics as shown by Raman spectroscopy, and uniform particle size distribution observed by electron microscopy of the barium titanate nanoparticles provide a foundation for subsequent piezoelectric properties, which are consistent with the piezoelectric properties after polarization. Regarding consistency, the uniformity of the piezoelectric nanoparticles in the ink forms the underlying basis. The barium titanate nanoparticles prepared by the hydrothermal method have uniform size and high crystallinity, achieving high dispersion and stable suspension in the ink system. This avoids particle agglomeration and phase separation from the source, ensuring absolute uniformity of the functional phase in spatial distribution.

[0060] Encapsulated barium titanate nanoparticle assemblies with dynamic assembly processes: Maintaining a completely liquid PDMS substrate during the printing process allows for complete penetration and encapsulation of the deposited barium titanate ink. A corresponding structural diagram is shown below. Figure 3 As shown in a, the PDMS substrate, after a brief period of uncured state reaching a weak viscoelastic state, only allows partial embedding of barium titanate ink, thus forming a semi-embedded piezoelectric structure. A schematic diagram of this semi-embedded state is shown in Figure a. Figure 3 As shown in b in the diagram. The PDMS substrate, cured before the printing process, completely prevents the penetration and embedding of barium titanate ink; under these conditions, the barium titanate ink adheres to the surface of the cured substrate only through physical adsorption. A schematic diagram of the relevant structure is shown in [reference needed]. Figure 3 c. Furthermore, the cross-section of this type of piezoelectric structure exhibits a tightly stacked, embedded configuration, characterized by a tight bond between the barium titanate linear layer and the PDMS substrate without significant interfacial gaps. The cross-sectional morphology of this structure is as follows: Figure 3 As shown in d in the figure. Accordingly, the cross-section of this type of semi-embedded structure shows that the barium titanate linear layer protrudes slightly from the PDMS substrate surface, which is a typical morphological feature of the semi-embedded state. Figure 3 (e). Furthermore, the cross-section of this type of surface-attached structure exhibits a flat, thin-layer configuration, which is significantly different from embedded and semi-embedded structures, as shown in the cross-sectional view below. Figure 3As shown in f, the core factor leading to such significant structural differences among the three embedded piezoelectric structures lies in the interfacial interaction between barium titanate ink and the PDMS substrate. These interfacial interactions directly affect the ink's penetration and embedding behavior on the substrate surface. For the fully liquid PDMS substrate, its excellent fluidity effectively supports complete embedding of the barium titanate ink while ensuring tight stacking of the piezoelectric phase within the substrate. However, for the short-term uncured PDMS substrate, its inherent viscoelasticity limits the penetration depth of the barium titanate ink, thereby reducing the degree of ink embedding and forming a semi-embedded structure. In contrast, the cured PDMS substrate completely loses its ability to encapsulate barium titanate ink, which further leads to significant differences in the embedding state and structural morphology of the barium titanate linear structure in the three embedded piezoelectric structures.

[0061] Figure 4 The effects of different printing parameters on the morphology of printed lines were demonstrated. Experimental results show that the line width can be precisely controlled by adjusting the printing pressure and printing speed. Specifically, as the printing pressure increases (Fig. df), the line width gradually increases; while as the printing speed increases (Fig. gi), the line width gradually decreases. Furthermore, the influence of nozzle height on line morphology was also investigated (Fig. ac). Through comparative analysis, we determined the optimal printing process parameters, ensuring the uniformity and continuity of the prepared lines, laying the foundation for the stable printing of subsequent large-scale array structures.

[0062] The flexibility of piezoelectric structures: Figure 5 The results of output stability tests of this device under cyclic mechanical loads under different experimental conditions are presented. Figure 5 (a) and Figure 5 (b) The performance degradation trend of non-embedded and semi-embedded devices after 10,000 cycles is presented. In contrast, Figure 5 (a) The embedded device maintains a stable output voltage after 1,000,000 cycles, demonstrating its excellent mechanical durability and long-term operating capability. Comparative analysis confirmed the process optimization potential of this structure and determined the key parameter range for achieving optimal cycle stability, providing important reliability data for the technology to move from the laboratory to practical applications.

[0063] Polarization control of piezoelectric properties: The macroscopic morphology of the flexible piezoelectric structure before and after polarization is shown in the figure. The structure in both states maintains a regular arrangement and uniform line spacing, indicating that the polarization process does not destroy the regularity and stability of the substrate structure. Figure 6 a and Figure 6 (b) The XRD pattern clearly shows the difference in crystal structure before and after polarization. Figure 6 c in Figure 6 In Figure d): The XRD peaks before polarization are wider and have lower intensity, while the characteristic peaks after polarization become sharper and have significantly increased intensity. This indicates that polarization treatment effectively increases the tetragonal phase content and crystal order of barium titanate, and further enhances the spontaneous polarization intensity. The voltage stability curve below directly shows the performance difference before and after polarization treatment. The voltage response curve before polarization can capture deformation signals, but there are large signal fluctuations and insufficient peak-to-valley consistency. Figure 6 In contrast, the polarized voltage curve has a more regular shape and a higher peak value, indicating that the material has excellent durability. Furthermore, the repeatability and stability of the signal are significantly improved after multiple tensile and bending tests. Figure 6 (f) In this process, polarization treatment induces the directional alignment of ferroelectric domains in barium titanate grains through an external electric field, while oil bath heating reduces the domain rotation resistance, making it easier for the domains to align with the external electric field and thus enhancing the spontaneous polarization effect. Simultaneously, PDMS molecular chains are adsorbed onto the barium titanate surface through van der Waals forces, which inhibits particle agglomeration and structural cracking, and maintains a uniform distribution of the piezoelectric phase during electromechanical coupling, ultimately improving the voltage stability and output performance of the structure under repeated deformation. Therefore, after polarization treatment, while retaining the regularity and stability of the substrate structure, the tetragonal phase content and crystal order of barium titanate are effectively increased, and the spontaneous polarization intensity is further enhanced. Furthermore, the voltage curve after polarization is more regular in shape and has a higher peak value. After multiple tensile and bending tests, the repeatability and stability of the signal are significantly improved. This indicates that the consistency of piezoelectric performance after nanoparticle polarization is closely related to the consistency of piezoelectric performance after assembly structure polarization, and the dual uniformity of material distribution and spatial structure is effectively transmitted to the consistency of piezoelectric performance regulated by polarization.

[0064] Figure 7 This demonstrates the feasibility of controlling device output performance by altering the printed pattern design. Experiments were conducted comparing single-unit ( Figure 7 a) One-dimensional array ( Figure 7 b) and two-dimensional grid array ( Figure 7 c) Electrical responses of three different structures. The results show that the output voltage increases significantly with increasing pattern complexity and effective active area, rising from approximately 3V for a single-line structure to approximately 35V for a grid array. Figure 7 (df ​​in the text). This conclusion confirms the flexibility and scalability of this printing process in realizing high-performance flexible electronic devices, and provides an effective strategy for customizing devices with specific electrical properties according to different application requirements.

[0065] Figure 8This diagram illustrates the polarization process and piezoelectric working mechanism of a flexible piezoelectric structure. The upper part shows the process by which the electric dipoles (blue arrows) of the barium titanate particles inside the material change from a random and disordered state to a highly ordered arrangement along a specific direction before and after polarization. The lower part shows the process by which a piezoelectric potential difference (separation of positive and negative charges) is generated on the upper and lower surfaces of the polarized flexible structure when it is subjected to external force (F) and undergoes bending deformation, thereby outputting current in the external circuit. The enlarged view below reveals the fundamental physical mechanism by which the deformation of the BaTiO3 cell (perovskite structure) at the microscopic level before and after the force causes the center of positive and negative charges to shift.

[0066] Figure 9 Five flexible linear devices of different lengths were characterized under specific pressure excitations of 0.1 N (Figure I) and 1 N (Figure II). Experimental results show that each device exhibits highly consistent dynamic response characteristics, with response times typically distributed within a very small range of 0.426 s to 0.489 s. This high degree of performance stability under different pressure loads strongly confirms the extremely high uniformity of the fabrication process and the robustness of the devices operating in complex mechanical environments. Furthermore, the insensitivity of electrical properties to geometric dimensions further reveals that this device possesses excellent design flexibility, performance stability, and is a high-performance component unaffected by manufacturing errors or pressure fluctuations.

[0067] Figure 10 The system demonstrates the dynamic charging characteristics of five piezoelectric arrays composed of five single-wire structures under pressure excitation of 0.1N (Figure I) and 1N (Figure II). Experimental results clearly show that the array structure significantly improves the device's output voltage, achieving output voltages exceeding 14V under a 1N load. Notably, each array unit exhibits extremely high consistency under the same pressure, with response times consistently remaining within the range of 0.628s to 0.678s. This result strongly demonstrates the superior uniformity and reliability of this fabrication process in multi-unit integration and large-area fabrication. Furthermore, the output characteristics of this array structure exhibit good stability and predictability under different loads, providing strong support for the practical application of this flexible piezoelectric structure in precision sensing and high-density energy conversion. This packaged piezoelectric assembly structure, combining flexibility and consistency, yielded a voltage of 51.2 pC·N. -1 Its piezoelectric coefficient is 1.87 V·kPa. -1 It exhibits linear sensitivity, an ultra-fine micro-pressure detection limit of 0.01 Pa, and a fast response time of 0.426 s.

[0068] Regulation of piezoelectric properties: The figure shows a detailed schematic diagram of the linear flexible piezoelectric structure, clearly presenting the structural configuration and distribution characteristics of the piezoelectric structure in the linear flexible mode. Figure 11 (a) When the structure is bent along the printing direction, its deformation is specifically transmitted along a straight line; this directional deformation induces ordered lattice distortion in the barium titanate nanoparticles, and the piezoelectric effect further transforms this lattice distortion into regular multi-peak voltage pulses—each bending action corresponds to a clearly distinguishable signal peak, directly reflecting the stable electromechanical conversion performance of this linear flexible piezoelectric structure under a specific bending direction. Figure 11 (b) When the structure is bent perpendicular to the printing direction, its internal stress distribution changes significantly compared to bending along the printing direction; this change in stress distribution causes the voltage waveform to exhibit irregular multi-peak characteristics, with significant fluctuations in both peak value and signal frequency. This irregular voltage response fully demonstrates the anisotropic sensing characteristics of the linear flexible piezoelectric structure, which is also a key feature that distinguishes this structure from other designs. Figure 11 (c) The figure shows a schematic diagram of the grid-like flexible piezoelectric structure, intuitively demonstrating the grid intersection structure characteristics and the distribution pattern of the piezoelectric elements in this flexible system. Figure 11 (d) When the flexible piezoelectric structure is bent along the printing direction, its voltage response exhibits a regular multi-peak pulse signal; this response mode is similar to the response characteristics of a linear structure when bent in the same direction, indicating that the electromechanical conversion mechanism has a certain consistency under this specific bending condition. Figure 11 (e) When the flexible mesh structure is bent perpendicular to the printing direction, the voltage curve still maintains a uniform multi-peak waveform without significant signal distortion. Figure 11 (f) This phenomenon demonstrates that the mesh structure design can effectively disperse the stress generated during bending. This effective stress dispersion allows deformation in different directions to be efficiently converted into electrical signals, further proving that the mesh flexible piezoelectric structure has superior isotropic sensing capability compared to linear flexible structures. Furthermore, this study further verified the influence of structural design on piezoelectric performance through tensile testing of the flexible piezoelectric structure; the relevant test results are shown in (f). Figure 27As shown. These supplementary experiments provide additional experimental evidence for the conclusions obtained from the bending test. For the linear flexible piezoelectric structure, under tension, its linear structure concentrates and transmits stress along a straight direction. This directional stress transmission endows the lattice distortion of barium titanate nanoparticles with obvious directionality, which is consistent with the stress transmission mechanism of the linear structure. Therefore, the electrical signal output of the linear flexible piezoelectric structure under tension is consistent with the response observed along the printing direction in the bending test, further confirming the directional characteristics of the linear structure in stress transmission and electromechanical conversion. For the grid flexible piezoelectric structure, its grid structure can uniformly disperse the stress generated during the stretching process, thereby avoiding the local stress concentration phenomenon that may occur in the linear structure. This uniform stress dispersion makes the lattice distortion of barium titanate nanoparticles more uniform and stable. Therefore, the electrical signal output of the grid flexible piezoelectric structure remains stable during the stretching process without significant fluctuations, further confirming the unique advantages of the grid structure in stress regulation and piezoelectric performance stability. In summary, the strip design of the linear structure imparts a clear directionality to stress transmission, resulting in significant differences in electrical response under different bending directions. In contrast, the cross-grid structure of the mesh structure breaks the directionality of stress transmission, enabling the structure to maintain stable electromechanical conversion efficiency under bending in all directions. Furthermore, the interfacial van der Waals forces between the components of the piezoelectric structure ensure the stable distribution of barium titanate nanoparticles during repeated bending cycles, effectively guaranteeing the long-term durability of the structure's piezoelectric properties.

[0069] Force sensing capability of piezoelectric structures Figure 12 This study demonstrates the feasibility of controlling device output performance through optimized printing pattern design. Experiments compared the electrical responses of three different structures: a single unit (Fig. a), a one-dimensional parallel array (Fig. b), and a two-dimensional cross-grid array (Fig. c). Results show that with increasing pattern complexity and effective active area, the device output voltage exhibits a significant increasing trend, with the peak output voltage rising from approximately 2.5V for a single linear array to over 30V for the two-dimensional grid array (Fig. I). Simultaneously, all three structures demonstrate excellent linear sensing characteristics within the 0-1 kPa pressure range (Fig. II) and possess sub-second fast response capabilities (Fig. III). This conclusion confirms the flexibility and scalability of this printing process in realizing high-performance flexible electronic devices, providing an effective structural design strategy for customizing sensing sensitivity and electrical performance according to different application scenarios.

[0070] Figure 13This study demonstrates the feasibility of controlling device output performance through optimized printing pattern design. Experiments compared the electrical responses of two different structures: a single unit (Figure a) and a one-dimensional parallel array (Figure b). The results show that with increasing pattern complexity and effective active area, the device output voltage exhibits a significant increasing trend, with the peak output voltage rising from approximately 3V for a single linear array to around 15V for a one-dimensional parallel array (Figure I). Simultaneously, both structures demonstrate good linear sensing characteristics within the 0-1 kPa pressure range and possess rapid dynamic response capabilities (Figure II). This conclusion confirms the flexibility and scalability of this printing process in realizing high-performance flexible electronic devices, providing an effective structural design strategy for customizing electrical performance according to different application scenarios.

[0071] Example 2 In this embodiment, the piezoelectric nanoparticle assembly structure prepared in Example 1 is applied to human force sensing and other applications.

[0072] Spinal test: To systematically investigate the ability of piezoelectric nanoparticle assemblies to transmit force in complex skeletons, piezoelectric nanoparticle assemblies were precisely deployed on the spinal nodes of the test subjects' backs. The electrical signals generated by pressure were then tested in upright, relaxed, or slouching postures.

[0073] Tactile perception test: The piezoelectric nanoparticle assembly structure is cut open to be used as a tentacle. By applying the same pressure and sliding motion on the same plane and non-flat plane, the dynamic tactile perception is reflected by electrical signals.

[0074] Vibration test: For accurate identification of multi-directional vibration, ten branches of the piezoelectric nanoparticle assembly structure are suspended and subjected to tilt, positive phase and bottom forces, and the electrical signals excited by each branch are received using a digital source meter.

[0075] Fabrication process and multi-scenario application framework of piezoelectric mesh structures based on direct writing printing technology, such as... Figure 1 As shown in a, the initial core steps of the entire process include preparing barium titanate nanoparticles using matching technology and formulating these nanoparticles into printable barium titanate ink with excellent rheological properties. Figure 25 The process comprises five consecutive steps: (1) coating liquid polydimethylsiloxane (PDMS) onto a polyethylene terephthalate (PET) substrate; (2) forming a uniform PDMS structure layer by spin coating; and (3) depositing barium titanate ink onto the liquid PDMS substrate using direct writing printing technology to construct a linear piezoelectric structure. Figure 26(4) A grid-like piezoelectric structure is prepared by secondary orientation rotation and direct writing printing of the linear piezoelectric structure; (5) The PET substrate is peeled off to obtain the preliminary piezoelectric structure. Specifically, PDMS prepared in proportion is coated onto the PET substrate by spin coating to ensure the uniformity and viscoelastic compatibility of the substrate; then, barium titanate ink is deposited on the uncured PDMS surface along a predetermined path by direct writing printing technology, and the combination of piezoelectric phases is achieved by utilizing the liquid properties of PDMS. Finally, the PET substrate is peeled off after curing treatment, and the grid structure is constructed by secondary printing to obtain a complete piezoelectric grid structure. This invention demonstrates the application of this structure in sensing ( Figure 1 (b) also demonstrates the application of this structure in shape perception ( Figure 1 (c) This invention describes the application of this structure in the power supply of micro-devices, and the method for implementing channel signal detection after array integration ( Figure 1 (d in the text)

[0076] Application of encapsulated piezoelectric nanoparticle components in using machine learning to analyze human motion: In spinal motion monitoring, piezoelectric structures were attached to the upper, middle, and lower parts of the spine, corresponding to the upper, middle, and lower parts of the voltage data graph. The mechanical response of the spine was then tested in five motion states: standing still, walking slowly, jogging, running quickly, and jumping. Figure 14(ad in the text). Each subgraph records voltage changes during three repeated motion processes, revealing the differentiated characteristics of spinal deformation under different motion intensities and verifying the good repeatability of the signal. When the human body is at rest, the spine bears a stable load, and the voltage signal output by the structure fluctuates very little. During slow walking, the spine undergoes periodic small-amplitude deformation with gait, thus generating regular low-amplitude voltage pulses. When the motion intensity increases to jogging, running, and jumping, the deformation amplitude and frequency of the spine increase significantly, which simultaneously increases the peak voltage and signal frequency output by the structure. There are significant differences in the voltage response amplitude at the upper, middle, and lower positions of the spine, reflecting the mechanical response characteristics of each segment of the spine under different motion states. The polarization direction of barium titanate grains couples with the stress direction during motion, inducing lattice distortion through electromechanical effects; ion displacement disrupts electroneutrality and generates a potential difference, thereby outputting a detectable voltage signal. When the motion stops, the spine returns to a stable state, and the voltage signal falls back to the baseline level. This study compared and analyzed the monitoring performance under different motion states. The results showed that the flexible piezoelectric structure maintained a stable voltage output at all locations on the spine and exhibited high signal consistency with no significant attenuation during three repeated motions. Therefore, this flexible piezoelectric structure not only possesses excellent flexibility and sensitivity but also achieves stable and repeatable electrical signal output in complex human motion scenarios, demonstrating significant application prospects in fields such as human motion tracking and spinal health assessment.

[0077] Piezoelectric nanoparticle assemblies have shown great application potential in the field of health monitoring, bringing development opportunities to wearable electronics and the medical and health sectors. Firstly, by attaching piezoelectric nanoparticle assemblies to a chair, a multi-point array layout can cover the upper, middle, and lower regions of the human back, corresponding to the thoracic, lumbar, and coccygeal vertebrae. When a person leans back in the chair, the pressure and contact area at different points will cause the piezoelectric nanoparticle assemblies to generate different electrical signal responses, enabling the capture of subtle differences in posture. For example... Figure 15 As shown in Figure a, when the human body is in an upright, relaxed, or slumped posture, there are significant differences in the local pressure and deformation exerted on the piezoelectric nanoparticle assembly structure by different parts of the spine, which converts these mechanical energies into electrical signals. Figure 15 (bd in the text). To achieve accurate recognition of spinal movements, deep feature extraction and machine learning processing were performed on the acquired electrical signals. We attached this flexible mesh piezoelectric structure to the skin of the human throat, demonstrating its application in voice recognition. Figure 15(e, f in the original text). When the wearer sequentially utters vowels and consonants (l, m, n), the vibrations of the laryngeal muscles and vocal cords are transmitted to the flexible mesh piezoelectric structure, causing lattice distortion in the barium titanate nanoparticles. The piezoelectric effect converts this mechanical deformation into a recognizable voltage signal. As shown in the corresponding waveforms, vowel signals exhibit higher peak voltages and wider pulse widths, while consonants produce shorter, more compact pulses. The significant differences in peak value, frequency, and duration provide a quantifiable electrical basis for speech recognition and articulation monitoring. The flexible PDMS substrate ensures a close fit to the skin, and the van der Waals forces at the interface maintain the uniform distribution of the barium titanate nanoparticles, thus guaranteeing high sensitivity and repeatability during the sound perception process. Figure 15 As shown in h, t-SNE feature clustering analysis is used to reduce the dimensionality of high-dimensional signal data. The electrical signals generated by three different spinal postures form three independent and clearly defined red, blue, and yellow regions in the two-dimensional feature space, demonstrating that the electrical signals captured by the piezoelectric nanoparticle assembly structure possess extremely high data discriminability and separability. Based on this, a classification algorithm is used for pattern recognition of the signal features, such as... Figure 15 As shown in i, the model's predictions for the three standard spinal movements perfectly match the actual categories, achieving 100% precision and recall. This result verifies that piezoelectric nanoparticle assembly structures, combined with machine learning, can efficiently and faithfully perform intelligent perception and accurate classification of human spinal movements.

[0078] Application of encapsulated piezoelectric nanoparticle assemblies in machine learning-based detection of irregular surface morphology (touch sensing): By placing piezoelectric nanoparticle assemblies as tactile devices on substrates of different materials and shapes, static and dynamic contact stresses can be monitored in various spatial scenarios. These piezoelectric nanoparticle assemblies can play a crucial role in tactile perception and shape recognition applications. For example... Figure 17 As shown in Figure a, when piezoelectric nanoparticle assembly structures are placed on the same plane and subjected to the same pressure, they can output a stable voltage. The piezoelectric nanoparticle assembly structures can accurately sense the uniform pressure state of a flat surface. For example... Figure 17 As shown in b, when a piezoelectric nanoparticle assembly is placed under different stepped, uneven surfaces, the stress response caused by changes in spatial height and contact posture is observed. The piezoelectric nanoparticle assembly can sensitively capture the contact pressure differences caused by local undulations in the substrate and convert them into voltage output. Due to its excellent flexibility, the piezoelectric nanoparticle assembly can perfectly adapt to macroscopic surfaces of different shapes and undulations, and can continuously and stably acquire changes in spatial morphology and interfacial pressure under various flat or uneven pressure conditions.

[0079] By applying piezoelectric nanoparticle assemblies as tactile devices to various target objects, surface morphology can be identified in diverse spatial scenarios. Piezoelectric nanoparticle assemblies can play a crucial role in multidimensional tactile sensing applications. For example... Figure 18 As shown in Figure a, when piezoelectric nanoparticle assemblies are attached to the same plane, they are subjected to uniform pressure, allowing for the observation of a single-peak voltage change. This structure can accurately sense the contact state of a flat surface. Figure 18 As shown in b, when the piezoelectric nanoparticle assembly structure is placed under different stepped non-flat surfaces, it can sensitively capture the multidimensional stress changes caused by the surface height differences, exhibiting a multi-peak voltage output. Due to its excellent flexibility, the piezoelectric nanoparticle assembly structure can perfectly adapt to surfaces with different undulations, and can continuously and stably capture changes in spatial morphology and texture under various contact conditions.

[0080] A modular tentacle array was constructed based on a flexible piezoelectric structure manufactured using direct-write printing technology. This array supports both independent and integrated sensing modes and provides a structural foundation for multidimensional tactile detection. Figure 28 The static tactile sensation of this structure attached to the paper base plane is as follows: Figure 19 As shown in a, when the structure lightly touches or presses the paper-based surface, the mechanical stress generated by the contact induces lattice distortion in the barium titanate nanoparticles. This distortion is converted into a clear single-peak voltage pulse through the piezoelectric effect. This effect also exhibits excellent response characteristics on the metal substrate surface, with a more significant peak voltage, demonstrating stability. Figure 29 Furthermore, it exhibited good repeatability and material response differences under different substrate adaptation conditions. Additionally, the tentacles-sensing test of the flexible piezoelectric structure on the same plane (…) Figure 30 a) This further validates the stable single-peak voltage output, enabling accurate signal capture even under no-load conditions. This study demonstrates the dynamic tactile response generated when the structure slides on a paper-based plane: stress changes resulting from the continuous interaction between the structure and the substrate are converted into unique voltage pulse signals, and the signals generated by repeated sliding are highly consistent, confirming the reliability of dynamic tactile perception. Figure 19 (b) It is worth noting that the flexible piezoelectric structure exhibits a significant signal enhancement effect in tactile perception on different spatial planes. Figure 30b). When the structure contacts different planes of an uneven carrier like a tentacle, the voltage peak value increases significantly compared to the voltage under the same plane condition. The waveform accurately reflects the spatial differences in the contact posture, thus effectively distinguishing between two-dimensional plane and three-dimensional space tactile perception. The structure exhibits complex tactile perception characteristics when sliding on uneven surfaces: surface undulations cause differentiated deformations in the structure, resulting in a multi-peaked and irregular voltage waveform; the changes in peak and valley values ​​precisely correspond to the positions of surface protrusions and depressions, clearly reproducing the surface microstructure, demonstrating the structure's high-resolution ability to capture complex surface textures. Figure 19 c). The t-SNE visualization of the sensor signal characteristics clearly shows the clustering separation effect among the three different morphological categories ( Figure 19 (e). The confusion matrix for the three-class action classification task shows that the model achieved 100% prediction accuracy when distinguishing between the three different forms. Figure 19 (d in the text). Furthermore, this study also achieved quantitative sensing of pressure intensity (…). Figure 31 (ab) When the flexible piezoelectric structure is subjected to pressure on the same plane, the output voltage maintains a stable pulse amplitude, demonstrating uniform pressure response characteristics. When gradient pressure is applied to different planes, the voltage peak increases stepwise with increasing pressure, thus achieving advanced functions from basic tactile recognition to precise pressure quantification. The core of this contact-deformation-electric signal conversion mechanism stems from the dual advantages of the flexible piezoelectric structure: the flexibility of the PDMS substrate allows it to closely conform to different surface contours, while the tetragonal phase of barium titanate nanoparticles efficiently converts mechanical stress into quantifiable electrical signals. Interfacial van der Waals forces ensure structural stability under dynamic contact conditions. The differences in voltage response under different tactile scenarios, substrate materials, spatial dimensions, and pressure intensities not only highlight the application potential of this structure in intelligent robot tactile perception, flexible electronic skin, and surface morphology detection, but also provide new technical support for the development of tactile interaction technology.

[0081] Application of encapsulated piezoelectric nanoparticle assemblies in machine learning-based sound signal recognition (vibration sensing): Through a series of sound production and acoustic perception experiments ( Figure 20 We systematically evaluated the sensing performance of the flexible mesh piezoelectric structure. Subsequently, we placed the flexible mesh piezoelectric structure in front of a speaker playing sound from 30 dB to 60 dB (30 dB, 40 dB, 50 dB, and 60 dB) and evaluated its response to sound pressure levels. Figure 20(a) As sound intensity increases, the voltage peak rises synchronously: the signal at 30 dB is weak and flat, while the signal amplitude at 60 dB increases significantly and the waveform becomes denser, clearly reflecting the positive correlation between sound decibel level and electrical output. The grid structure enhances the sound wave capture effect, while the tetragonal piezoelectric properties of barium titanate ensure efficient sound-to-electric conversion. To further verify its ability to distinguish environmental sound events, we conducted two typical action tests on the flexible grid piezoelectric structure: opening a door and opening a window (…). Figure 20 b and Figure 20 (c) Each action produces a unique voltage waveform with characteristic differences in amplitude, duration, and response pattern, which can serve as discriminative features for classification. We then employ the t-SNE feature clustering method ( Figure 20 The separability of these voltage signals is visualized using d) in the diagram. The "door open" and "window open" categories form independent and non-overlapping clusters in the feature space, indicating that the structure has strong discriminative power. The corresponding confusion matrix ( Figure 20 (e) confirms the perfect classification accuracy, demonstrating the reliable performance of the flexible mesh piezoelectric structure in identifying different environmental sound events. Notably, the polarization state of the barium titanate nanoparticles plays a crucial role in the sensing performance of the flexible mesh piezoelectric structure. Unpolarized barium titanate, although a tetragonal ferroelectric material, has randomly oriented polarization domains, leading to mutual cancellation of macroscopic polarization effects; under acoustic vibration, this only generates weak and unstable charge signals with low discrimination capability. In contrast, after polarization by an external electric field, the polarization domains align uniformly, thereby activating the macroscopic piezoelectric effect. The lattice deformation induced by sound waves generates constructively superimposed charges, significantly improving signal strength, sensitivity, and stability. This explains the clearer and more consistent correlation between sound intensity and electrical signal output observed in the polarized barium titanate sample.

[0082] like Figure 21 As shown, ten spatially distributed branches of a piezoelectric nanoparticle assembly structure are assembled to sense different vibrations. Because the piezoelectric nanoparticle assembly structure has multiple branches, it generates different electrical signals for the same vibration. By distinguishing the electrical signals of different vibrations, accurate identification of multi-directional vibrations can be achieved.

[0083] like Figure 22 As shown in Figure ac, different wind directions are used to simulate the forces of different vibrations. When a piezoelectric nanoparticle assembly structure with ten spatially distributed branches is subjected to tilted, positive, and bottom-direction forces, the wind force causes different degrees of bending deformation in each branch due to the spatial arrangement differences, thus outputting electrical signals with significantly different characteristics. Figure 22As shown in d, t-SNE dimensionality reduction analysis maps the five-channel signal features into a two-dimensional space, demonstrating that vibrations in different directions form well-defined clusters in the feature space. Figure 22 As shown in e, by using machine learning classification algorithms to train and infer the extracted features, the specific direction of external excitation can be automatically determined with extremely high accuracy, ultimately achieving highly sensitive capture and accurate intelligent recognition of multi-dimensional and complex directional vibrations.

[0084] In summary, this invention applies the piezoelectric nanoparticle assembly structure formed by direct writing printing inside a liquid PDMS substrate to the fields of human health monitoring, sports equipment assistance, and multidimensional tactile sensing. The response mechanism and electrical signal output characteristics of the piezoelectric nanoparticle assembly structure under different mechanical interaction scenarios were systematically investigated, and the experimental results are as follows: (1) The piezoelectric nanoparticle assembly structure can accurately sense the continuous movements of the spine. This piezoelectric nanoparticle assembly structure senses the dynamic stress applied to the spine and converts it into characteristic electrical signals. Combined with t-SNE dimensionality reduction analysis and machine learning classification algorithm, it achieves highly sensitive capture and accurate intelligent recognition of various spinal movements.

[0085] (2) In tactile perception mode, the piezoelectric nanoparticle assembly structure, acting as a tentacle, demonstrates a comprehensive capability of both statically accurate identification and dynamically high-resolution tracking of complex surface morphologies. In static contact scenarios, the piezoelectric nanoparticle assembly structure can keenly capture and distinguish the differences in local multidimensional stress distribution between two-dimensional flat surfaces and three-dimensional stepped undulating surfaces by outputting a stable and consistent single-peak voltage or a significantly enhanced multi-peak voltage. During dynamic sliding interaction, when the piezoelectric nanoparticle assembly structure slides over paper-based or uneven surfaces, it can sensitively convert the continuous bending and rebound stress caused by alternating surface textures and continuous friction into multiple repetitive and highly consistent continuous multi-peak pulse signals. This piezoelectric nanoparticle assembly structure, acting as a tentacle, combined with t-SNE dimensionality reduction analysis and machine learning classification algorithms, achieves accurate identification of surface materials and textures.

[0086] (3) In vibration sensing scenarios, piezoelectric nanoparticle assemblies are used to distinguish between throat sound vibrations and acoustic sound vibrations. The ten-branched piezoelectric nanoparticle assembly can differentiate the local spatial multidimensional stress distribution differences caused by different vibrations. When the piezoelectric nanoparticle assembly is subjected to tilting, positive phase, and bottom-direction forces, wind forces cause different degrees of bending deformation in each branch, resulting in the output of electrical signals with significantly different characteristics. Combining t-SNE dimensionality reduction feature clustering analysis with machine learning classification algorithms, accurate and intelligent identification of environmental vibrations is successfully achieved.

[0087] Example 3 In this embodiment, a piezoelectric nanoparticle assembly structure was prepared by printing BaTiO3 nanoparticle ink onto a liquid uncured PDMS substrate, as described in Example 1. The self-powered performance of the piezoelectric nanoparticle assembly structure was investigated. It was found that due to its unique self-encapsulation characteristics and the presence of the piezoelectric effect, this structure can efficiently collect mechanical energy generated by human movement and convert it into electrical energy, exhibiting excellent energy conversion behavior. This makes it suitable for self-powered applications such as streetlight warnings, decorative lighting, and smart displays. Simultaneously, a multilayer perceptron (MLP) neural network architecture was combined to achieve controllable high-voltage prediction and output. High-voltage discharge phenomena at the tip were observed by accurately predicting and cascading multiple piezoelectric nanoparticle assemblies. Pale purple and pale blue sparks were excited in humid air and ethanol gas, respectively, enabling the ionization, differentiation, and rapid screening of specific chemical gases. This piezoelectric nanoparticle assembly structure can efficiently collect mechanical energy generated by the environment and human movement, showing great application potential in self-powered sensing, wearable electronic devices, and smart displays. It provides a feasible approach for developing a new generation of customized, highly durable, non-traditional self-powered power generation devices.

[0088] Intelligent Display Testing: Even the smallest everyday movements of the human body can serve as excellent mechanical power sources for self-powered systems. Taking intelligent pharmaceutical packaging as an example, a piezoelectric nanoparticle assembly structure is integrated onto the surface of the medicine bottle label. When a finger presses this structure, the applied fingertip pressure is instantly converted into electrical energy. This electrical energy is directly used to light up different numbers of small bulbs, thus displaying different information intuitively and clearly.

[0089] Lighting Test: Addressing the battery life bottleneck of personal portable electronic devices, this study further developed a self-powered lighting solution based on large-scale human movement. Piezoelectric nanoparticle assemblies were attached to the grip of handheld devices and the soles of wearable shoes, with small light bulbs connected to both ends of the assemblies. During hand gripping or daily walking, the gripping force generated by muscle contraction and the stepping force from body weight are efficiently transmitted to the structure, illuminating the light bulbs.

[0090] Controllable High Voltage Output Testing: To achieve accurate prediction and quantitative application of piezoelectric output performance, an MLP neural network model was introduced. Using experimental test data under different surface pressures and different numbers of printed layers as the training set, the MLP model can accurately fit and predict the extreme high voltage output trend under specific operating conditions. Guided by the model, multiple optimized piezoelectric structures were connected in series, and an external force with a potential sufficient to break down the gas was applied. High-voltage discharge was observed in air and ethanol environments. Piezoelectric nanoparticle assembly structure harvests mechanical energy from human movement: like Figure 23As shown in Figure a, the dosage information on the packaging label is displayed by pressing light bulbs with different numbers on them, which can solve the confusion experienced by elderly people with declining vision when taking different doses of medication. The piezoelectric nanoparticle assembly structure can effectively harvest the mechanical energy generated by pressing the finger and convert it into electrical energy. This electrical energy is displayed as key information about the drug dosage through small light bulbs with different numbers. This smart display simplifies complex information into easily perceptible signals, lowering the barrier for users to understand and receive dosage requirements. For handheld lighting devices, batteries need to be replaced or recharged regularly when they run out. Figure 23 As shown in b, the piezoelectric nanoparticle assembly structure can effectively harvest the mechanical energy generated by finger gripping and convert it into electrical energy, thereby supporting long-term lighting effects. For handheld ambient lighting devices, the piezoelectric nanoparticle assembly structure can harvest and convert the mechanical energy generated during the gripping process into electrical energy, creating an ambient lighting effect with a miniaturized, lightweight, self-powered unit. In outdoor scenarios, wearable shoe lighting devices with piezoelectric nanoparticle assembly structures can harvest and convert mechanical energy through walking to meet lighting needs, representing a portable, self-sufficient solution.

[0091] Controllable high voltage output achieved through piezoelectric nanoparticle assembly structure: like Figure 24 As shown in Figure a, the MLP neural network architecture enables the qualitative harvesting of mechanical energy from the piezoelectric nanoparticle assembly structure and its conversion into electrical energy, thereby generating a high-voltage output at the tip. The MLP neural network architecture can be used to predict the voltage generated by multiple piezoelectric nanoparticle assemblies. For example... Figure 24 As shown in b, the dataset contains the voltages of piezoelectric nanoparticle assemblies with a length of 4 cm and a diameter of 40 μm under different numbers of printed lines and different surface pressures. Using an MLP neural network architecture, the voltages of piezoelectric nanoparticle assemblies with different numbers of printed lines under different surface pressures can be predicted. The grip force of a human hand is approximately 500 N; using an MLP neural network, the voltage generated by piezoelectric nanoparticle assemblies with different numbers of printed lines under this surface pressure can be predicted. Figure 24 As shown in Figure c, the precise high voltage generated by the piezoelectric nanoparticle assembly structure can be used for high-voltage discharge. Connecting multiple piezoelectric nanoparticle assemblies in series makes it easier to trigger tip discharge, producing a pale purple spark in humid air. Ethanol, a common small organic molecule model, can be ionized through the tip discharge of multiple series-connected piezoelectric nanoparticle assemblies, producing a pale blue spark. The different colored sparks generated by the piezoelectric nanoparticle assembly structure can be used to distinguish the presence of certain substances in the air, which can be used for rapid screening of industrial leaks, indoor air quality testing, and hazardous gas early warning in the safety field.

[0092] In this embodiment, a piezoelectric nanoparticle assembly structure was prepared by direct writing BaTiO3 nanoparticle ink onto a liquid uncured PDMS substrate. The self-powered energy collection performance of this structure in terms of environmental and human mechanical energy was investigated. Furthermore, the controllable prediction of high-voltage output and its application in gas detection were explored using an MLP neural network. The experimental results are as follows: (1) The piezoelectric nanoparticle assembly structure can efficiently harvest mechanical energy generated by human movement and has been successfully applied to intelligent interactive and portable lighting devices. Under the action of finger pressing, the structure harvests mechanical energy and converts it into electrical energy, which can light up small light bulbs with different digital information. As a smart packaging label, it can intuitively display the drug dosage, effectively reducing the threshold for elderly people with declining vision to receive complex dosage information. At the same time, under the action of hand grip or daily walking, the structure can directly provide long-term power support for handheld ambient lighting devices or wearable shoe lighting devices, realizing lightweight, portable energy self-sufficiency.

[0093] (2) Accurate prediction and quantitative control of the output voltage of the piezoelectric nanoparticle assembly structure were achieved by combining MLP neural networks. With increasing surface pressure, the mechanical deformation applied to the structure increases, triggering a stronger piezoelectric effect, resulting in a significantly higher predicted output voltage. Based on the precise high voltage generated by the piezoelectric nanoparticle assembly structure, the detection of tip discharge and ionization of specific chemical gases was successfully realized. The assembly structure corresponding to the voltage can be accurately predicted using the MLP neural network. Connecting multiple piezoelectric nanoparticle assemblies in series can easily trigger tip discharge. In humid air, tip discharge produces a pale purple spark; while in an environment using ethanol as an organic small molecule model, the discharge ionizes the ethanol, producing a pale blue spark. This phenomenon of generating different colored sparks through self-powered operation can effectively distinguish the presence of specific substances in the air.

[0094] In summary, the piezoelectric nanoparticle assembly structure applied to this flexible BaTiO3 nanoparticle assembly structure can efficiently convert human motion energy into electrical energy, demonstrating great application potential in lighting, early warning, decoration, display, controllable voltage output, and hazardous gas early warning.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for preparing a highly flexible and uniform piezoelectric nanoparticle assembly, characterized in that, Includes the following steps: The solution-direct writing printing technology is used to directly write barium titanate onto an uncured elastic substrate. After printing, it is cured by heating and polarization to obtain the final product.

2. The preparation method according to claim 2, characterized in that, The conditions for heat curing are: heat curing at 80~90℃ for 30~50 minutes; Preferably, the polarization treatment conditions are: oil bath heating polarization treatment at 70~90℃ and 400~500V.

3. The preparation method according to claim 1, characterized in that, The specific steps of the solution direct writing printing include: the liquid viscoelastic substrate is fixed on the printing station, barium titanate ink is injected into the ink cartridge, and direct writing printing is performed. Preferably, the specific process of direct write printing is as follows: the ink cartridge uses a printing needle to perform direct write printing according to a preset program with set pressure, speed and height; Preferably, a printing needle with a diameter of 100-200 micrometers is used; the printing pressure is 5-20 psi, preferably 7-10 psi; the printing speed is 30-60 mm / s; and the printing height is 0-1.5 mm. Preferably, during printing, the pattern design adopts a single unit, a one-dimensional array, or a two-dimensional grid array; the one-dimensional array is a piezoelectric array composed of multiple single-line structures, which includes radial or ring-shaped branching structures with parallel tentacles; Alternatively, barium titanate ink is prepared by using barium titanate nanoparticles as the functional phase precursor and a mixture of dimethyl sulfoxide and ethylene glycol as the composite dispersion solvent. The two are mixed to obtain a barium titanate precursor dispersion, which is then subjected to ultrasonic treatment. Preferably, the mass fraction of barium titanate nanoparticles is 30-50 wt%. Preferably, in the composite dispersion solvent, the volume ratio of dimethyl sulfoxide to ethylene glycol is (0.9~1.1):(0.9~1.1); Alternatively, the liquid viscoelastic substrate may be a viscoelastic polydimethylsiloxane substrate; Preferably, the viscoelastic polydimethylsiloxane substrate is prepared by mixing a polydimethylsiloxane precursor with a curing agent to prepare liquid PDMS, and then spin-coating the liquid PDMS onto the surface of a pretreated PET substrate. Preferably, the polydimethylsiloxane precursor and the curing agent are mixed and reacted in a vacuum environment for 0.5 to 5 hours; Preferably, the mass ratio of polydimethylsiloxane precursor to curing agent is (5~15):1; Preferably, the specific spin coating conditions are as follows: 280~320 rpm for 20~40 seconds, 380~420 rpm for 20~40 seconds, 480~520 rpm for 20~40 seconds, 580~620 rpm for 20~40 seconds, 680~720 rpm for 20~40 seconds, and 980~1020 rpm for 20~40 seconds.

4. A highly flexible and uniform piezoelectric nanoparticle assembly, characterized in that, It is prepared by the preparation method described in any one of claims 1 to 3.

5. The piezoelectric nanoparticle assembly according to claim 4, characterized in that, Barium titanate is completely encapsulated in an elastic substrate.

6. A multi-dimensional force sensing device, characterized in that, Including the highly flexible and consistent piezoelectric nanoparticle assembly as described in claim 5.

7. The application of the highly flexible and uniform piezoelectric nanoparticle assembly of claim 5, or the highly flexible and uniform piezoelectric nanoparticle assembly prepared by the preparation method of any one of claims 1 to 3, or the multidimensional force sensing device of claim 6, in multidimensional and multimodal force sensing.

8. The application according to claim 7, characterized in that, The multidimensional force sensing includes analyzing human motion, detecting irregular surface morphology, and using machine learning to identify sound signals; Preferably, the multimodal force sensing applications include self-powered sensors, wearable electronic devices, and smart displays.

9. The application according to claim 7, characterized in that, When applying sensing, either independent sensing or integrated sensing can be used.

10. A method for multi-dimensional force sensing, characterized in that, The highly flexible and consistent piezoelectric nanoparticle assembly as described in claim 4 or the multidimensional force sensing device as described in claim 5 are employed.

Images