4d printed components based on ald and methods of making and smart packaged devices

Abstract

This invention discloses an ALD-based 4D printed component, its fabrication method, and a smart packaging device. The component includes a three-dimensional programmable substrate and an atomic layer deposition (ALD) functional nanolayer formed on at least a portion of the surface of the three-dimensional programmable substrate. The thickness, material, regional distribution, and / or stacking sequence of the ALD functional nanolayer are configured to actively adjust the deformation path of the three-dimensional programmable substrate under external stimuli, and convert the external stimuli or the deformation path into readable electrical, dielectric, optical, thermal, or chemical functional signals. It participates in 4D deformation programming through thermal expansion mismatch, residual stress, surface adsorption, photothermal conversion, or dielectric modulation. This approach enables the ALD film layer to simultaneously serve as a deformation programming layer and a signal conversion layer, making it suitable for flexible sensing, smart chip packaging, micro-actuators, soft robots, and structural health monitoring.

Description

Technical Field

[0001] This invention relates to the fields of additive manufacturing, atomic layer deposition, smart materials, flexible electronics, advanced packaging, soft robotics, and 4D printing. Specifically, it relates to an ALD-based 4D printed component and its preparation method, as well as a smart packaging device. In particular, it relates to a 4D printed component, its preparation method, and its application that introduces an atomic layer deposition (ALD) functional nanolayer into a 3D printed programmable substrate and utilizes the material, thickness, regional distribution, and stacking sequence of the nanolayer to actively participate in stimulus response, deformation path programming, and functional signal readout. Background Technology

[0002] Traditional 3D printing primarily addresses geometric shaping issues, with the functionality of components typically relying on the printing material itself. When functionalities such as conductivity, dielectric properties, humidity sensitivity, photothermal properties, barrier properties, or interface monitoring are required, the conventional approach is to directly add functional fillers to the printing ink, or to form a surface layer after printing through methods such as dip coating, spraying, electroplating, chemical plating, or sputtering. However, directly adding functional fillers to the printing material can easily lead to increased viscosity, uneven curing, nozzle clogging, decreased mechanical properties, and reduced deformation reversibility. Conventional surface coating methods also struggle to form uniformly thick, interface-stable, and precisely controllable functional films within the inner walls of pores, microbeam sidewalls, origami hinges, lattice rods, and porous networks.

[0003] 4D printing utilizes shape memory polymers (SMPs), hydrogels, liquid crystal elastomers, or multi-material structures to achieve reversible bending, rolling, folding, expansion, contraction, or aperture adjustment, showing potential applications in micro-actuators, flexible sensing, soft robotics, and smart packaging. However, the deformation paths of existing 4D printed components are mostly determined by the matrix material composition, geometry, and printing direction, and functional signal readout often relies on additionally attached strain gauges, electrodes, or sensors. This approach suffers from problems such as unstable interface bonding, difficulty in covering complex three-dimensional surfaces, asynchrony between local response and signal readout, and challenges in miniaturization and integration.

[0004] ALDs possess self-limiting surface reactions, nanoscale thickness control, and excellent conformal coverage capabilities, enabling the formation of continuous thin films on the outer surfaces, inner walls, and curved surfaces of complex 3D structures. When using ALDs for 3D printed structures, they are typically used as protective coatings, barrier layers, surface hardening layers, or chemically modified layers, primarily focusing on wear resistance, corrosion resistance, water and oxygen barrier properties, dielectric insulation, or surface energy modulation. Such approaches do not incorporate the thickness, material, regional distribution, residual stress, and stacking sequence of the ALD film as 4D deformation programming variables, nor do they synchronously convert the deformation process into readable electrical, dielectric, optical, or chemical signals.

[0005] The simple combination of "4D printed components plus surface deposition layers" can be understood as a conventional approach, as coatings for protection or modification are already publicly available in various fields. The real technical challenges are: how to transform nanoscale ALD films from passive coverings into functional layers capable of actively controlling macroscopic deformation paths, response thresholds, cyclic stability, and functional signal output; how to enable stable cross-scale coupling within local hinges, microbeams, pore walls, and encapsulation interfaces without sacrificing the flexibility of the 4D printed substrate; and how to make this cross-scale coupling predictable and repeatable through process parameters and calibration models.

[0006] Therefore, it is necessary to propose a new 4D printed component based on ALD functionalization and its preparation method, so that the ALD film layer can simultaneously serve as a deformation programming layer, a stimulus response conversion layer and a functional signal readout layer, and form a synergy with the three-dimensional programmable matrix at the structural, material, interface and signal levels. Summary of the Invention

[0007] The purpose of this invention is to address the problems existing in the prior art by providing an ALD-based 4D printed component, its preparation method, and an intelligent packaging device.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In a first aspect, an ALD-based 4D printed component is provided, comprising a three-dimensional programmable substrate and an atomic layer deposition functional nanolayer formed on at least a portion of the surface of the three-dimensional programmable substrate; the thickness, material, regional distribution, and / or stacking order of the atomic layer deposition functional nanolayer are configured to actively adjust the deformation path of the three-dimensional programmable substrate under external stimuli, and convert the external stimuli or the deformation path into readable electrical, dielectric, optical, thermal, or chemical functional signals.

[0009] Furthermore, the thickness of the atomic layer deposition functional nanolayer is from 0.5 nm to 500 nm, preferably from 2 nm to 100 nm; at least two regions on the three-dimensional programmable substrate have different atomic layer deposition cycle numbers, so that the two regions have different film thicknesses, different residual stresses, or different interface stiffnesses.

[0010] Furthermore, the atomic layer deposition functional nanolayer is selectively formed on at least one of the following locations of the three-dimensional programmable substrate: one side surface, two sides surface, local hinge region, microbeam upper surface, channel inner wall, origami unit crease region, lattice rod surface, or encapsulation interface region.

[0011] Furthermore, the atomic layer deposition functional nanolayer includes one or more of Al2O3, TiO2, ZnO, HfO2, ZrO2, SiO2, SnO2, VO2, V2O5, NiO, CuO, RuO2, IrO2, Pt, Pd, TiN, MoS2, WS2, or nanolayers thereof.

[0012] Among them, Al2O3, HfO2 or ZrO2 layers are used to form dielectric monitoring or blocking signal layers, ZnO, SnO2 or NiO layers are used to form humidity or gas adsorption signal layers, Pt, Pd or TiN layers are used to form photothermal, conductive or resistive signal layers, and MoS2 or WS2 layers are used to form strain-sensitive or photoelectric response layers.

[0013] Furthermore, the three-dimensional programmable matrix includes shape memory polymers, hydrogels, elastomers, liquid crystal elastomers, photocurable resins, polyimides, thermoplastic polymers, ceramic / polymer composites, metal / polymer composites, or carbon-based composites.

[0014] Furthermore, the three-dimensional programmable substrate and the atomic layer deposition functional nanolayer together form a stimulus-response coupling unit, which is a composite interface region between the two. Within this composite interface region, the material, thickness, regional distribution, and stacking sequence of the atomic layer deposition functional nanolayer have a preset mapping relationship with the geometric configuration, material properties, and deformation mode of the three-dimensional programmable substrate.

[0015] The preset mapping relationships include mappings of bending angle and resistance change rate, curvature and capacitance change rate, aperture change rate and dielectric constant change, interface stress and leakage current change, curl radius and optical transmittance change, or humidity stimulation and impedance change.

[0016] Furthermore, the atomic layer deposited functional nanolayer modulates the deformation path or functional signal of the three-dimensional programmable substrate through at least one mechanism among thermal expansion coefficient mismatch, film residual stress, surface adsorption, photothermal conversion, interface dielectric modulation, or conductive path reconstruction.

[0017] Furthermore, it also includes electrodes, flexible interconnects, contact pads, encapsulation leads, or wireless readout units connected to the atomic layer deposition functional nanolayer for reading the functional signals.

[0018] Furthermore, the atomic layer deposition functional nanolayer is a multilayer stacked structure, which includes at least two types of sublayers: a deformable programming sublayer, a signal conversion sublayer, and a barrier protection sublayer, and the thickness of each sublayer is controlled by different atomic layer deposition cycles.

[0019] Furthermore, when the 4D printed component is used for smart chip packaging, the three-dimensional programmable substrate is a packaging buffer frame, a serpentine buffer structure, an origami buffer structure, or a locally porous buffer structure, and the atomic layer deposition functional nanolayer is used to simultaneously perform water and oxygen barrier, interface stress adjustment, and packaging health status signal readout. When the 4D printed component is used for flexible pressure sensing, the three-dimensional programmable matrix is ​​a porous elastomer or hydrogel structure, and the atomic layer deposition functional nanolayer conformally covers the inner wall of the pores and generates changes in capacitance, resistance or impedance signals under external pressure, humidity or gas stimulation.

[0020] Furthermore, the 4D printed components are used in flexible sensors, smart chip packaging, micro actuators, soft robots, wearable electronics, environmental monitoring, medical microdevices, or structural health monitoring.

[0021] Secondly, a method for preparing 4D printed components based on ALD is provided, the method comprising the following steps: Design target stimuli, target deformation, and target functional signals; 3D printing of three-dimensional programmable substrates; The three-dimensional programmable substrate is surface activated or regionally passivated. Atomic layer deposition is used to form a functional nanolayer in a predetermined area. The deposition temperature of the atomic layer deposition is 25°C to 150°C, the precursor pulse time is 0.01s to 10s, the reactant pulse time is 0.01s to 10s, the purge time is 1s to 120s, and the film thickness, residual stress or signal sensitivity are controlled by the number of deposition cycles. Forming electrodes and signal readout terminals; The relationship between external stimuli, deformation, and functional signals is trained and calibrated.

[0022] Furthermore, the training calibration includes collecting external stimuli S(t), deformation δ(t), film thickness h, film material M, deposition region R, substrate geometric parameters G, and functional signal F(t), and establishing a response mapping model for predicting deformation path and functional signal; the response mapping model is a multivariate regression model, Gaussian process regression model, random forest model, gradient boosting tree model, neural network model, or physically constrained neural network model, and its output includes bending angle, curvature, response time, recovery time, signal sensitivity, cycle drift rate, or failure risk; the training data of the response mapping model comes from finite element simulation, composite beam theoretical calculation, atomic layer deposition process records, deformation test, electrical test, humidity test, thermal cycling test, or packaging reliability test, and at least a portion of the samples are used to verify the correspondence between film parameters and deformation path and functional signal.

[0023] The surface activation includes oxygen plasma treatment, argon plasma treatment, ultraviolet ozone treatment, hydroxylation treatment, or chemical grafting treatment; the regional passivation includes forming a photoresist mask, a metal mask, a polymer suppression layer, a fluorosilane passivation layer, or a local sacrificial protection layer.

[0024] Furthermore, after the training calibration, a cyclic stability evaluation step is included, wherein the cyclic stability evaluation records the deformation recovery rate, functional signal drift rate, film cracking state, interface peeling state, and electrode contact resistance after at least 10 to 10,000 stimulation cycles.

[0025] Thirdly, a smart packaging device is provided, including a chip, a packaging substrate, a flexible interconnect structure, and the aforementioned ALD-based 4D printed component, wherein the 4D printed component is disposed at the edge of the chip, the corner of the packaging substrate, the periphery of the flexible interconnect, or the interface stress concentration area.

[0026] Compared with existing technologies, the beneficial effects of this invention are: 1. This ALD-based 4D printed component elevates the atomic layer deposition film from a traditional protective coating, barrier coating, or surface modification layer to a functional nanolayer that actively participates in 4D deformation programming and functional signal conversion; 2. Utilizing the conformal deposition capability of ALD, continuous functional films can be formed on the inner walls of holes, hinges, microbeams, and lattice rods of complex 4D printed structures, solving the defect that traditional coatings cannot cover complex three-dimensional surfaces; 3. Utilizing the nanoscale control capability of ALD film thickness, macroscopic deformation paths, response thresholds, and signal sensitivity can be achieved by changing the number of deposition cycles. 4. By utilizing the dielectric, conductive, humidity-sensitive, photothermal, and barrier properties of different ALD materials, multi-region, multi-stimulus, and multi-signal outputs can be achieved on the same 4D printed component; 5. The ALD film layer serves as both a deformation programming layer and a signal conversion layer, reducing reliance on external sensors or complex functional inks and lowering the difficulty of directly printing complex functional materials; 6. Through response training calibration and cyclic stability evaluation, a repeatable correspondence between film parameters, deformation paths, and functional signals can be established, improving the reliability of components in flexible sensing, smart packaging, micro-actuators, soft robots, and wearable electronics. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the structure of a 4D printed component based on ALD according to the present invention; Figure 2 This is a schematic diagram illustrating the 4D deformation programming of the ALD film thickness, material, and regional distribution in this invention. Figure 3 This is a schematic diagram of the conformal deposition mechanism of ALD on complex 3D printed structures according to the present invention; Figure 4 This is a schematic diagram illustrating the physical mechanism by which the ALD membrane layer of this invention participates in the stimulus response; Figure 5 This is a schematic diagram of the fabrication method of ALD-based functionalized 4D printed components according to the present invention. Figure 6 This is an exemplary comparative diagram showing the effect of ALD film thickness on deformation and functional signals according to the present invention; In the figure: 100, three-dimensional programmable substrate; 200, atomic layer deposition functional nanolayer; 300, stimulus-response coupling unit. Detailed Implementation

[0028] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Example 1: A 4D printed component based on ALD is provided, such as... Figure 1 As shown, it includes a three-dimensional programmable substrate 100 and an atomic layer deposition functional nanolayer 200 formed on at least a portion of the surface of the three-dimensional programmable substrate 100; the thickness, material, regional distribution and / or stacking order of the atomic layer deposition functional nanolayer 200 are configured to actively adjust the deformation path of the three-dimensional programmable substrate under external stimuli, and convert the external stimuli or the deformation path into readable electrical, dielectric, optical, thermal or chemical functional signals.

[0030] This 4D printed component elevates the ALD (Atomic Layer Deposition) film from a traditional protective coating, barrier coating, or surface modification layer to a functional nanolayer that actively participates in 4D deformation programming and functional signal conversion. The component uses a three-dimensional programmable substrate as the macroscopic deformation carrier and the ALD functional nanolayer as the interface regulation and signal conversion carrier. The stimulus-response coupling unit 300 is formed through the composite interface region between the two, enabling the nanoscale film parameters to affect macroscopic bending, folding, pore size changes, curl radius, or micro-deformation of the encapsulation interface, and converting deformation or external stimuli into resistance, capacitance, impedance, leakage current, optical transmittance, dielectric constant, or chemical response signals.

[0031] The 4D printing referred to in this invention refers to the process of creating an initial geometric structure through 3D printing, and then enabling the component to undergo designable changes in shape, performance, or signal in response to external stimuli such as temperature, humidity, light, electric field, gas, mechanical load, or chemical environment over time.

[0032] Further, the atomic layer deposition functional nanolayer 200 includes one or more of Al2O3, TiO2, ZnO, HfO2, ZrO2, SiO2, SnO2, VO2, V2O5, NiO, CuO, RuO2, IrO2, Pt, Pd, TiN, MoS2, WS2, or nanolayers thereof.

[0033] Among them, Al2O3, HfO2 or ZrO2 layers are used to form dielectric monitoring or blocking signal layers, ZnO, SnO2 or NiO layers are used to form humidity or gas adsorption signal layers, Pt, Pd or TiN layers are used to form photothermal, conductive or resistive signal layers, and MoS2 or WS2 layers are used to form strain-sensitive or photoelectric response layers.

[0034] The thickness of the atomic layer deposition functional nanolayer 200 is 0.5 nm to 500 nm, preferably 2 nm to 100 nm; at least two regions on the three-dimensional programmable substrate have different atomic layer deposition cycle numbers, so that the two regions have different film thicknesses, different residual stresses, or different interface stiffnesses.

[0035] For Al2O3 films deposited using trimethylaluminum and water as precursors, the film thickness can be continuously varied in the range of several nanometers to tens of nanometers by adjusting the number of cycles; similar nanoscale thickness adjustment can also be achieved for ZnO, TiO2 or HfO2 films using the corresponding precursor systems.

[0036] Furthermore, the atomic layer deposition functional nanolayer 200 is selectively formed on at least one of the following locations of the three-dimensional programmable substrate 100: one side surface, two sides surface, local hinge region, microbeam upper surface, channel inner wall, origami unit crease region, lattice rod surface, or encapsulation interface region.

[0037] Furthermore, the three-dimensional programmable substrate 100 includes shape memory polymers, hydrogels, elastomers, liquid crystal elastomers, photocurable resins, polyimides, thermoplastic polymers, ceramic / polymer composites, metal / polymer composites, or carbon-based composites. The substrate structure can be a thin film, cantilever beam, double-layer beam, origami unit, lattice unit, porous network, microchannel structure, helical structure, rollable sheet, or encapsulated buffer frame. This substrate is responsible for providing the macroscopic geometry, initial compliance, and basic deformation path.

[0038] Furthermore, the three-dimensional programmable substrate 100 and the atomic layer deposition functional nanolayer 200 together form a stimulus-response coupling unit 300, which is a composite interface region between the two. Within this composite interface region, the material, thickness, regional distribution, and stacking order of the atomic layer deposition functional nanolayer have a preset mapping relationship with the geometric configuration, material properties, and deformation mode of the three-dimensional programmable substrate.

[0039] The preset mapping relationships include mappings of bending angle and resistance change rate, curvature and capacitance change rate, aperture change rate and dielectric constant change, interface stress and leakage current change, curl radius and optical transmittance change, or humidity stimulation and impedance change.

[0040] The stimulus-response coupling unit 300 is not a standalone electronic circuit, nor is it an abstract functional concept. Rather, it is a composite interface region formed by a three-dimensional programmable substrate and an ALD functional nanolayer. Within this region, there is a pre-defined mapping relationship between the material, thickness, regional distribution, and stacking sequence of the ALD film and the geometry, material properties, and deformation mode of the substrate. For example, the thermal expansion mismatch and residual stress between a single-sided Al2O3 film and a shape memory polymer microbeam can alter the curvature; a ZnO film conformally covering the inner wall of the pores can change the conductivity upon moisture adsorption, while changes in pore size alter the effective contact area; and an HfO2 or Al2O3 layer covering the encapsulation buffer frame can generate changes in capacitance, leakage current, or impedance signals upon micro-deformation of the interface.

[0041] External stimuli can be temperature, humidity, light, electric field, gas concentration, pH value, mechanical tensile, compression or bending load; deformation can be bending angle, curvature, displacement, aperture change rate, curl radius or volume change rate; functional signals can be resistance, capacitance, current, leakage current, dielectric constant, optical transmittance, infrared reflectance or impedance.

[0042] In this embodiment, the material, thickness, regional distribution, and stacking sequence of the atomic layer deposition functional nanolayer 200 are used to adjust the response threshold, bending direction, bending curvature, response speed, recovery time, signal sensitivity, and cycle stability. For example, depositing a high-modulus oxide thin layer on one side of the substrate can induce asymmetric strain distribution under thermal stimulation; conformally depositing a humidity-sensitive film layer on the inner wall of the pores can simultaneously change the pore adsorption characteristics and electrical signals; depositing a dielectric layer on the encapsulation buffer frame can enable the stress relief structure to also have capacitance or leakage current monitoring functions.

[0043] In this embodiment, the physical mechanisms for adjusting the deformation path or functional signal of the three-dimensional programmable substrate 100 include, but are not limited to, thermal expansion coefficient mismatch, residual stress in the film layer, interface stiffness modulation, surface adsorption, dielectric modulation, conductive path reconstruction, and photothermal conversion. Taking a thermally responsive composite beam as an example, the linear expansion coefficient of the ALD oxide film layer is typically lower than that of the shape memory polymer or hydrogel matrix. When the temperature rises, there is a difference between the expansion of the matrix and the expansion of the film layer, resulting in a bending tendency of the double-layer beam. As the film thickness increases, the bending constraint, residual stress, and interfacial shear effect of the film layer are enhanced, causing measurable changes in the bending curvature and recovery path. Taking a humidity-responsive porous structure as an example, the ALD-ZnO or SnO2 film layer undergoes surface adsorption on the inner wall of the pores and changes its conductivity. The hydrogel or elastomer matrix simultaneously undergoes expansion or pore size changes. The two responses are coupled at the same pore interface, causing the deformation and electrical signal to evolve synchronously.

[0044] Furthermore, the 4D printed component also includes electrodes, flexible interconnects, contact pads, encapsulation leads, or wireless readout units connected to the atomic layer deposition functional nanolayer 200 for reading the functional signals.

[0045] Furthermore, the atomic layer deposition functional nanolayer 200 is a multilayer stacked structure, which includes at least two types of sublayers: a deformable programming sublayer, a signal conversion sublayer, and a barrier protection sublayer, and the thickness of each sublayer is controlled by different atomic layer deposition cycles.

[0046] Furthermore, when the 4D printed component is used for smart chip packaging, the three-dimensional programmable substrate 100 is a packaging buffer frame, a serpentine buffer structure, an origami buffer structure, or a locally porous buffer structure, and the atomic layer deposition functional nanolayer 200 is used to simultaneously perform water and oxygen barrier, interface stress adjustment, and packaging health status signal readout. When the 4D printed component is used for flexible pressure sensing, the three-dimensional programmable substrate 100 is a porous elastomer or hydrogel structure, and the atomic layer deposition functional nanolayer 200 conformally covers the inner wall of the pores and generates changes in capacitance, resistance or impedance signals under external pressure, humidity or gas stimulation.

[0047] Example 2: This example uses shape memory polymer microbeams as an example to further explain in detail how the ALD film in Example 1 controls macroscopic deformation through nanoscale parameters (the physical mechanism of ALD film participating in 4D deformation programming).

[0048] Combination Figure 2 and Figure 4As shown, the thickness of the three-dimensional programmable substrate ranges from 50 μm to 1000 μm, and the thickness of the ALD (Atomic Layer Deposition) film ranges from 5 nm to 50 nm. Although the film thickness is much smaller than the substrate thickness, the ALD film typically has a high elastic modulus, a low coefficient of thermal expansion, and adjustable residual stress. When the temperature changes or the substrate undergoes shape recovery, the thermal expansion mismatch and interfacial constraints between the film and the substrate will change the neutral layer position and curvature evolution process of the composite beam, thereby causing a measurable change in the macroscopic bending angle.

[0049] Taking Al2O3 film and shape memory polymer matrix as an example, the linear expansion coefficient of the oxide film can be significantly lower than that of the polymer matrix. The polymer matrix undergoes significant expansion or shape recovery during heating, while the film constrains its surface. If the film is located on one side of the microbeam, the strain on both sides of the composite beam is asymmetrical, resulting in directional bending. If the film is located on both sides with different thicknesses, the constraints on both sides are unbalanced, which can also form a controllable curvature. If the film is only located in the hinge region, the bending stiffness and recovery sequence at the hinge are preferentially adjusted, and the overall origami structure can be unfolded or folded in a predetermined order.

[0050] During the growth of ALD films, tensile or compressive stresses may be generated, which are related to deposition temperature, precursor system, reactants, purging time, and cycle number. As the number of deposition cycles increases, the film thickness and the stress accumulation effect within the film intensify, potentially altering the initial curvature, thermal response threshold, and cycle recovery path of the composite structure. By setting different cycle numbers in different regions, multiple bending units can be formed on the same substrate, causing them to exhibit different bending amplitudes or different response sequences under external stimuli.

[0051] Surface adsorption and functional signal conversion mechanisms are applicable to moisture-sensitive, gas-sensitive, or chemically responsive components. For example, after depositing ZnO, SnO2, or NiO films on the pore walls of 3D-printed hydrogels or porous elastomers, water or gas molecules enter the pores and adsorb on the film surface, changing the carrier concentration or interfacial charge state, thereby causing changes in resistance or impedance. Simultaneously, the substrate's hygroscopic or swelling leads to changes in pore size, diffusion channels, and effective contact area, so that the electrical signal reflects not only the adsorption behavior of the film material itself but also the deformation process of the 4D-printed substrate.

[0052] Photothermal and electrical conductivity conversion mechanisms are applicable to microactuators or optically driven components. For example, depositing Pt, TiN, or TiO2 / Pt stacks in local hinge regions can induce localized heating in these regions under light or electrical excitation. This localized heating further triggers softening of shape memory polymers, orientation changes in liquid crystal elastomers, or volume changes in hydrogels, thereby causing the component to bend, curl, or fold. Because ALDs can control the film thickness and deposition area, the photothermal conversion region can be precisely separated from the non-responsive region, avoiding indiscriminate heating of the entire component.

[0053] Example 3: The advantage of ALD lies in its ability to conformally deposit complex three-dimensional structures. To enable the film layer to participate in deformation programming rather than indiscriminately covering the entire component, it combines... Figure 3 As shown, this embodiment provides a region-selective ALD process.

[0054] First, the deposition area is determined based on the target deformation path, such as the upper surface of the microbeam, the origami hinge area, the inner wall of the channel, the edge of the encapsulation interface, or the local stress concentration area; then, the target area is surface activated, and the non-target area is masked or passivated; finally, the film deposition is completed in the low-temperature ALD reaction chamber.

[0055] In one masking process, after the 3D programmable substrate 100 is printed, it is first cleaned and then cured. A metal mask, flexible polyimide mask, or photoresist pattern is then applied to the area where deposition is not required. For planar or gently curved microbeams, photoresist spin coating, exposure, and development can be used to form the pattern; for structures with large curvature or porosity, a peelable polymer inhibition layer, a sprayed mask, or a locally applied blocking material can be used. After deposition, the mask is removed, leaving the ALD film primarily in the predetermined area.

[0056] In a surface passivation process, areas where deposition is not desired are coated with a thin layer of fluorosilane, alkylsilane, polydimethylsiloxane, or other nucleation inhibition layer to eliminate reactive hydroxyl groups or active sites on their surface. Areas where deposition is desired are treated with oxygen plasma, ultraviolet ozone, hydroxylation, or chemical grafting to increase nucleation density. Selective deposition patterns can be formed by the difference in nucleation rates between the activated and passivated regions. For structures where deposition is desired on the inner wall of the pores but not on the outer surface, a removable sealing layer can be applied to the outer surface, leaving only the pore entrance intact, allowing the precursor to preferentially enter and react on the inner wall of the pores.

[0057] Within a specific process window, Al₂O₃ films can use trimethylaluminum and water as the precursor system, with a deposition temperature of 60°C to 120°C, a trimethylaluminum pulse time of 0.02 s to 1 s, a nitrogen or argon purging time of 5 s to 60 s, a water pulse time of 0.02 s to 1 s, followed by a second purging for 5 s to 60 s. ZnO films can use a diethylzinc and water system, with a deposition temperature of 50°C to 120°C. TiO₂ films can use a titanium source with water or ozone, and Pt films can use a metal-organic precursor and a reducing reactive gas, preferably employing a lower temperature or plasma-enhanced process. The above parameters are not the only limitations; those skilled in the art can adjust them within this range according to the substrate's temperature resistance, the film material, and the target film thickness.

[0058] To avoid thermal deformation of the polymer matrix or dehydration of the hydrogel, low-temperature ALD, space-based ALD, or plasma-enhanced ALD are preferred. For SMP or hydrogels with lower temperature resistance, the deposition temperature can be controlled between 25°C and 80°C, and conformality can be improved by extending the precursor exposure time, reducing the purge flow rate, or using a segmented deposition method. For polyimide or high-temperature resins used for encapsulation, a process temperature of 80°C to 150°C can be used to improve film density.

[0059] Example 4: A method for preparing 4D printed components based on ALD is provided.

[0060] Combination Figure 5 As shown, the preparation method includes the following steps: S1: Design target stimuli, target deformation, and target functional signals according to application requirements; For example, a thermally responsive microactuator can be set to temperature or light, with the target deformation being the bending angle or curl radius, and the target signal being resistance or optical transmittance; a humidity sensing component can be set to relative humidity, with the target deformation being the aperture change rate, and the target signal being impedance or capacitance; a smart packaging component can be set to thermal cycling or mechanical bending, with the target deformation being the micro-displacement of the packaging buffer frame, and the target signal being leakage current, capacitance, or interface impedance.

[0061] S2: 3D printed three-dimensional programmable substrate; For macroscopic flexible structures, fused deposition modeling, digital light processing, stereolithography, or direct ink writing can be used; for microscale actuators, two-photon polymerization or high-precision projection photolithography can be used; for hydrogel or functional ink structures, extrusion direct writing printing can be used; after printing, depending on the material type, cleaning, post-curing, solvent exchange, heat treatment, drying, or pre-training are performed to ensure that the substrate has a stable initial shape and repeatable response.

[0062] S3: Perform surface activation or regional passivation on the three-dimensional programmable substrate; The areas where ALD films need to be deposited are treated with oxygen plasma, argon plasma, ultraviolet ozone, hydroxylation, or chemical grafting to improve nucleation uniformity; the areas where films do not need to be deposited are masked, selectively passivated, or locally coated with an inhibition layer to reduce the deposition rate; the purpose of this step is not simply to clean the surface, but to match the future film distribution with the deformation path of the component.

[0063] S4: Atom layer deposition functional nanolayers are formed in a predetermined area using atomic layer deposition; The treated substrate is placed in the ALD reaction chamber, and precursors, reactants and purge gases are alternately introduced to grow functional films layer by layer through self-limiting surface reactions. By changing the number of ALD cycles, deposition temperature, precursor exposure time, regional masking or selective surface activation methods, thickness gradients, material gradients or regional selective patterns can be formed. For multifunctional components, a dielectric layer can be deposited first, followed by local deposition of a humidity-sensitive layer or a photothermal layer to form a multilayer stacked structure. For example, the deposition temperature of the atomic layer deposition is 25°C to 150°C, the precursor pulse time is 0.01s to 10s, the reactant pulse time is 0.01s to 10s, the purging time is 1s to 120s, and the film thickness, residual stress or signal sensitivity are controlled by the number of deposition cycles.

[0064] S5: Forms the electrode and signal readout terminal; For reading resistance, capacitance, leakage current, or piezoresistive signals, metal electrodes, flexible interconnects, contact pads, or encapsulation leads can be formed on the atomic layer deposition functional nanolayer 200. Electrodes can be formed by evaporation, sputtering, screen printing, inkjet printing, conductive silver paste dotting, or transfer. To reduce the limitation of electrode deformation, electrodes can be constructed using serpentine traces, island-bridge structures, liquid metal interconnects, or conductive polymers.

[0065] S6: Training and calibrating the relationship between external stimuli, deformation, and functional signals; The deposited components are subjected to thermal cycling, humidity cycling, light cycling, mechanical cycling, atmosphere exposure, or encapsulation reliability tests. Stimulus quantity S(t), deformation δ(t), and functional signal F(t) are recorded. Deformation can be measured using image recognition, laser displacement gauges, digital image correlation, microscopy, or profilometers. The functional signal can be measured using a source meter, impedance analyzer, capacitance meter, spectrometer, or data acquisition module. Response threshold, response time, recovery time, sensitivity, hysteresis, drift, and cycling stability are obtained through calibration.

[0066] Furthermore, the training calibration includes collecting external stimuli S(t), deformation δ(t), film thickness h, film material M, deposition region R, substrate geometric parameters G, and functional signal F(t), and establishing a response mapping model for predicting deformation path and functional signal; the response mapping model is a multivariate regression model, Gaussian process regression model, random forest model, gradient boosting tree model, neural network model, or physically constrained neural network model, and its output includes bending angle, curvature, response time, recovery time, signal sensitivity, cycle drift rate, or failure risk; the training data of the response mapping model comes from finite element simulation, composite beam theoretical calculation, atomic layer deposition process records, deformation test, electrical test, humidity test, thermal cycling test, or packaging reliability test, and at least a portion of the samples are used to verify the correspondence between film parameters and deformation path and functional signal.

[0067] The surface activation includes oxygen plasma treatment, argon plasma treatment, ultraviolet ozone treatment, hydroxylation treatment, or chemical grafting treatment; the regional passivation includes forming a photoresist mask, a metal mask, a polymer suppression layer, a fluorosilane passivation layer, or a local sacrificial protection layer.

[0068] Furthermore, after the training calibration, a cyclic stability evaluation step is included, wherein the cyclic stability evaluation records the deformation recovery rate, functional signal drift rate, film cracking state, interface peeling state, and electrode contact resistance after at least 10 to 10,000 stimulation cycles.

[0069] Example 5: This example provides a response mapping model and machine learning calibration method to further illustrate the preparation method in Example 4.

[0070] To avoid black-boxing of machine learning features, this embodiment limits the input, output, data sources, and technical functions of the response mapping model. This model serves to determine ALD film parameters and the physical structure of 4D components. Its input features include at least film thickness h, film material M, deposition region R, stacking sequence L, substrate geometry G, substrate modulus E, substrate thickness t, external stimulus history S(t), deposition temperature TALD, and electrode arrangement parameters P. Output targets include bending angle θ, curvature κ, pore size change rate η, response time tr, recovery time trec, resistance change rate ΔR / R0, capacitance change rate ΔC / C0, signal drift rate D, and film cracking risk Crack.

[0071] Training data can consist of three sources: the first is theoretical or finite element simulation data, including simulations of composite beam bending, thermal expansion mismatch, interfacial stress, and pore size changes; the second is ALD process record data, including film material, cycle number, deposition temperature, pulse time, purge time, and film thickness test results; the third is sample test data, including bending angle, curvature, resistance, capacitance, impedance, response time, cycle drift, and failure state. For the small sample stage, 30 to 200 sets of samples and simulation data can be used for initial calibration; for the mass production stage, this can be expanded to 200 to 5000 sets of data.

[0072] If a physically constrained neural network is used in the model, constraints such as the monotonicity of the composite beam curvature, non-negativity of film thickness, non-negativity of signal sensitivity, maximum strain limit, and film cracking threshold can be added to the loss function. If Gaussian process regression is used, the prediction uncertainty can be output to indicate which film thickness or region combinations require additional experiments. Regardless of the model used, its function is to infer the ALD film cycle number, material combination, and deposition region from the target deformation and target signal, rather than performing general post-processing on the signal.

[0073] In a specific training process, the dataset is divided into training, validation, and test sets at 70%, 15%, and 15% respectively. Inputs include film thickness, material encoding, deposition region encoding, substrate thickness, stimulus amplitude, and electrode spacing. Outputs include bending angle, capacitance change rate, and signal drift rate after 100 cycles. After training, if the mean absolute error of the bending angle on the test set is less than 5°, the capacitance change rate error is less than 10%, and the drift rate error is less than 3 percentage points, the model is considered suitable for the next round of structural design. The recommended scheme output by the model still needs to be validated using at least one set of samples. The validation results are fed back into the dataset to achieve closed-loop optimization of design, manufacturing, and operation.

[0074] Example 6: To illustrate that the ALD film layer does indeed participate in 4D deformation programming and functional signal readout, this example provides a set of exemplary simulation or prototype verification data.

[0075] The test subject was a shape memory polymer microbeam with dimensions of 20 mm × 3 mm × 0.3 mm. An Al2O3 dielectric film was deposited on one side, and flexible electrodes were formed at both ends of the film. The thermal stimulation conditions were cycling from 25 °C to 70 °C for 60 seconds. Comparative samples included samples without an ALD film, samples with a 5 nm film, samples with a 20 nm film, and samples with a 50 nm film.

[0076] Test results are as follows Figure 6As shown in Table 1, under the same thermal stimulation conditions, the bending angle of the component without an ALD film is approximately 23°, the capacitance change rate is approximately 0.8%, and the signal drift rate after 100 cycles is approximately 18.0%; the bending angle of the 5nm Al2O3 film sample is approximately 31°, the capacitance change rate is approximately 6.2%, and the drift rate is approximately 7.8%; the bending angle of the 20nm sample is approximately 47°, the capacitance change rate is approximately 14.5%, and the drift rate is approximately 4.1%; and the bending angle of the 50nm sample is approximately 63°, the capacitance change rate is approximately 23.0%, and the drift rate is approximately 6.8%. These results indicate that changes in ALD film thickness not only alter the deformation amplitude but also synchronously change the functional signal output, suggesting that the film simultaneously serves as a deformation programming layer and a signal conversion layer.

[0077] Table 1: Exemplary effects of ALD film thickness on 4D deformation and functional signal

[0078] It should be noted that while a 50nm film layer achieves a higher bending angle and signal change rate, its thickness may increase the risk of interfacial strain and cracking. Therefore, the maximum film thickness is not preferred for all applications. For disposable or low-cycle sensing components requiring high sensitivity, a thicker film layer can be selected; for encapsulated monitoring or wearable components requiring long-term cycling stability, a film layer of around 20nm or a multilayer film stack can be chosen. This conclusion reflects the designability of the present invention, namely, achieving a balance between deformation amplitude, signal sensitivity, and cycling reliability through film thickness, material, and regional distribution.

[0079] Example 7: This example provides the application of the 4D printed components in Example 1 in smart chip packaging.

[0080] In smart chip packaging, the 3D programmable substrate 100 is a 3D-printed serpentine buffer frame, origami buffer frame, or locally porous buffer structure, disposed at the chip edge, around the flexible interconnect, at the corner of the packaging substrate, or in areas of concentrated thermomechanical stress. The atomic layer deposition functional nanolayer 200 is an Al2O3, HfO2, ZrO2, or Al2O3 / TiO2 nanolayer stack, covering the surface of the buffer frame and part of the packaging interface.

[0081] This film layer provides water and oxygen barrier and dielectric isolation, and reflects the health status of the package through changes in leakage current, capacitance, or interface impedance. During temperature cycling or mechanical bending, the 3D-printed buffer frame undergoes reversible micro-deformation, and the ALD film layer generates changes in dielectric or leakage current signals as the interface stress changes. Compared to simply attaching strain gauges to the package, the functional film layer in this embodiment is located directly on the buffer structure and the package interface, enabling closer monitoring of local stress changes at the failure source.

[0082] In a preferred structure, a porous buffer frame is provided at the corner of the package substrate. A 20nm to 80nm Al2O3 / HfO2 stack is deposited on the surface of the frame, and electrodes are provided at opposite ends. When the package warps during thermal cycling from -40°C to 125°C, the compression or stretching of the buffer frame causes changes in equivalent capacitance and leakage current. If the signal drift at a certain corner exceeds a threshold, it can be determined that there is a risk of interface debonding, crack initiation, or stress concentration in that area, thus providing a basis for package reliability monitoring.

[0083] Example 8: This example provides the application of the flexible pressure sensor in Example 1 in a multi-stimulus response.

[0084] In flexible pressure sensing, the 3D programmable substrate 100 is a 3D-printed porous elastomer whose pore structure undergoes reversible compression under external pressure. HfO2, Al2O3, or ZrO2 dielectric layers are deposited on the pore walls and upper and lower surfaces using ALD (Alternating Current Dielectric Deposition), and flexible electrodes are placed on the upper and lower surfaces. External pressure compresses the pores, causing changes in the effective dielectric constant, electrode spacing, and contact area, thus inducing changes in the capacitance signal. Because the ALD dielectric layer has a uniform thickness on the inner walls of the pores, it reduces signal drift caused by uneven distribution of traditional dip-coating materials within the porous structure.

[0085] In the multi-stimulus responsive component, the three-dimensional programmable substrate 100 can be divided into a humidity-sensitive region, a photothermal region, and a dielectric monitoring region. The humidity-sensitive region is deposited with a ZnO or SnO2 film to output humidity or gas adsorption signals; the photothermal region is deposited with a Pt or TiO2 / Pt film to absorb light energy and induce local bending; and the dielectric monitoring region is deposited with an Al2O3 or HfO2 film to output capacitance or leakage signals. These three regions are formed separately using masking, selective activation, or surface passivation methods, enabling the same 4D-printed component to produce different deformations and signal outputs under different stimuli.

[0086] In this embodiment of the soft robot micro-gripper, the micro-gripper finger is printed with a shape memory polymer, with a Pt / TiO2 photothermal layer deposited on the inner side and an Al2O3 dielectric layer deposited on the outer side. When illuminated, the inner photothermal layer locally heats up, triggering the gripper to bend, while the outer dielectric layer provides feedback on the gripping state through capacitance changes. Since both the driving and signal layers are formed using ALD (Al2O3 polymer-modified plastic film), the films can conformally cover the curved finger, avoiding the limitations imposed on microstructure movement by traditional patch-type sensors.

[0087] Example 9: This example demonstrates the cyclic stability, quality control, and mass production of 4D printed components.

[0088] To demonstrate that the atomic layer-deposited functional nanolayer 200 is not a typical protective coating, cyclic response testing of the components is necessary. For example, a humidity-responsive component can be cycled between 20%RH and 90%RH, recording changes in pore size, resistance, and recovery time. For a thermally-responsive component, the cycle can be performed between room temperature and the substrate's response temperature, recording changes in bending angle, capacitance, and recovery angle. For an encapsulated buffer component, temperature and bending cycles can be performed, recording changes in leakage current, capacitance, impedance, and contact resistance. If the deformation and functional signals maintain a stable mapping relationship after multiple cycles, it indicates that the ALD film participates in 4D response and signal readout.

[0089] After manufacturing, samples should undergo at least visual inspection, dimensional inspection, film thickness testing, critical function testing, and reliability testing. Film thickness can be confirmed using ellipsometry, cross-sectional electron microscopy, X-ray reflectance, or standard sheet calibration. Electrical connection structures should be tested for open circuits, short circuits, contact resistance, and insulation resistance. Structures with thermal management or photothermal conversion should be tested for temperature rise, temperature gradient, and thermal cycling stability. Structures with humidity-sensitive or gas-sensitive responses should be tested for response time, recovery time, hysteresis, and drift. Structures with encapsulation monitoring functions should be tested for thermal cycling, bending cycling, and interface signal changes.

[0090] Before mass production or sample delivery, key process parameters and test results for each sample can be recorded to form a traceable implementation record. This record may include substrate material batch, printing parameters, post-curing conditions, surface activation conditions, mask material, deposition temperature, precursor pulse time, purge time, deposition cycle number, film thickness, surface morphology, electrode size, initial signal, operating signal, abnormal signal, and failure location. With these records, those skilled in the art can directly implement the invention according to the specification, and can also make equivalent substitutions under different materials, sizes, or operating conditions without requiring a new inventive design.

[0091] The above-described embodiments of the present invention can be combined with each other. For example, a smart packaging component can simultaneously employ region-selective ALD, response mapping model, and cycle stability evaluation; a flexible pressure sensing component can simultaneously employ a porous substrate, dielectric ALD layer, and machine learning calibration; a micro actuator can simultaneously employ a photothermal ALD layer and a resistive readout layer. As long as the thickness, material, regional distribution, or stacking order of the ALD film layer participates in 4D deformation programming and outputs functional signals, it falls within the protection scope of the present invention.

[0092] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An ALD-based 4D printed component, characterized in that, It includes a three-dimensional programmable substrate and an atomic layer deposition functional nanolayer formed on at least a portion of the surface of the three-dimensional programmable substrate; the thickness, material, regional distribution and / or stacking sequence of the atomic layer deposition functional nanolayer are configured to actively adjust the deformation path of the three-dimensional programmable substrate under external stimuli, and convert the external stimuli or the deformation path into readable electrical, dielectric, optical, thermal or chemical functional signals.

2. The ALD-based 4D printed component according to claim 1, characterized in that, The thickness of the atomic layer deposition functional nanolayer is from 0.5 nm to 500 nm; at least two regions on the three-dimensional programmable substrate have different atomic layer deposition cycle numbers, so that the two regions have different film thicknesses, different residual stresses, or different interface stiffnesses.

3. The 4D printed component based on ALD according to claim 1, characterized in that, The three-dimensional programmable substrate and the atomic layer deposition functional nanolayer together form a stimulus-response coupling unit. The stimulus-response coupling unit is a composite interface region between the two. Within this composite interface region, the material, thickness, regional distribution, and stacking order of the atomic layer deposition functional nanolayer have a preset mapping relationship with the geometric configuration, material properties, and deformation mode of the three-dimensional programmable substrate.

4. The ALD-based 4D printed component according to claim 1, characterized in that, It also includes electrodes, flexible interconnects, contact pads, package leads, or wireless readout units connected to the atomic layer deposition functional nanolayer for reading the functional signals.

5. The ALD-based 4D printed component according to claim 1, characterized in that, The atomic layer deposition functional nanolayer is a multilayer stacked structure, which includes at least two types of sublayers: a deformable programming sublayer, a signal conversion sublayer, and a barrier protection sublayer. The thickness of each sublayer is controlled by different atomic layer deposition cycles.

6. The 4D printed component based on ALD according to claim 1, characterized in that, When the 4D printed component is used for smart chip packaging, the three-dimensional programmable substrate is a packaging buffer frame, a serpentine buffer structure, an origami buffer structure, or a locally porous buffer structure. The atomic layer deposition functional nanolayer is used to simultaneously perform water and oxygen barrier, interface stress adjustment, and readout of packaging health status signals. When the 4D printed component is used for flexible pressure sensing, the three-dimensional programmable matrix is ​​a porous elastomer or hydrogel structure, and the atomic layer deposition functional nanolayer conformally covers the inner wall of the pores and generates changes in capacitance, resistance or impedance signals under external pressure, humidity or gas stimulation.

7. The method for preparing 4D printed components based on ALD according to any one of claims 1 to 6, characterized in that, The preparation method includes the following steps: Design target stimuli, target deformation, and target functional signals; 3D printing of three-dimensional programmable substrates; The three-dimensional programmable substrate is surface activated or regionally passivated. Atomic layer deposition is used to form a functional nanolayer in a predetermined area. The deposition temperature of the atomic layer deposition is 25°C to 150°C, the precursor pulse time is 0.01s to 10s, the reactant pulse time is 0.01s to 10s, the purge time is 1s to 120s, and the film thickness, residual stress or signal sensitivity are controlled by the number of deposition cycles. Forming electrodes and signal readout terminals; The relationship between external stimuli, deformation, and functional signals is trained and calibrated.

8. The method for preparing 4D printed components based on ALD according to claim 7, characterized in that, The training calibration includes collecting external stimuli S(t), deformation δ(t), film thickness h, film material M, deposition region R, substrate geometric parameters G, and functional signal F(t), and establishing a response mapping model for predicting deformation path and functional signal; the response mapping model is a multiple regression model, Gaussian process regression model, random forest model, gradient boosting tree model, neural network model, or physically constrained neural network model, and its output includes bending angle, curvature, response time, recovery time, signal sensitivity, cycle drift rate, or failure risk; The training data for the response mapping model comes from finite element simulation, composite beam theoretical calculation, atomic layer deposition process records, deformation testing, electrical testing, humidity testing, thermal cycling testing, or packaging reliability testing, and at least a portion of the samples are used to verify the correspondence between film parameters and deformation paths and functional signals.

9. The method for preparing 4D printed components based on ALD according to claim 7, characterized in that, The training calibration is followed by a cyclic stability evaluation step, which records the deformation recovery rate, functional signal drift rate, film cracking state, interface peeling state, and electrode contact resistance after at least 10 to 10,000 stimulation cycles.

10. A smart packaging device, characterized in that, The invention includes a chip, a packaging substrate, a flexible interconnect structure, and an ALD-based 4D printed component as described in any one of claims 1 to 6, wherein the 4D printed component is disposed at the edge of the chip, the corner of the packaging substrate, the periphery of the flexible interconnect, or the area of ​​interface stress concentration.

Images