I. Core Research Summary

• Support-free aerosol 3D‑printed near‑solid metal truss

Relying on the balance between droplet surface tension and self-weight, the structure requires no sacrificial supports throughout its entire fabrication. With truss diameters of 30–60 μm and unit-cell sizes of 100–400 μm, it represents a nearly solid micron‑scale metallic lattice that is difficult to achieve with conventional manufacturing processes, featuring controllable hierarchical porosity of 15%–20% within the trusses.

• The dual-deformation mechanism can be freely programmed and controlled.

Octahedral and hexagonal lattices are dominated by bending deformation, exhibiting a broad, stable stress plateau and excellent energy-absorption characteristics. In contrast, lattices with vertical struts undergo layered buckling failure, resulting in periodic oscillations of the stress plateau; switching between these mechanical behaviors can be achieved simply by modifying the printing path in CAD.

• Establishing a novel semi-empirical mechanical model adapted to ordered point arrays.

Traditional foam models are unable to capture the periodic hardening/softening behavior of lattice‑based platforms. In this study, an oscillatory correction term is introduced to accurately account for the stress fluctuations arising from the layer‑by‑layer collapse of multi‑layer unit cells, resulting in excellent agreement between simulations and experiments.

• FEA finite element analysis completes deformation visualization and prediction.

A three-dimensional beam element model is employed to accurately reproduce the three-stage deformation behavior: the elastic region, the stress plateau, and densification. For the bending-type lattice, the simulation error is only 3.1%; for the buckling-type lattice, the layered collapse mechanism can be qualitatively reproduced, with only a minor deviation in the amplitude of stress oscillations.

• Manufacturing defects have a negligible impact on macroscopic mechanical behavior.

Local printing defects such as variations in truss thickness and irregularities in cell geometry do not significantly alter the overall stress–strain curve, resulting in high process robustness and making it well suited for rapid, large‑scale prototyping.

II. Complete Preparation Process for 3D Micro‑Lattice Aerosols

(1) Complete Preparation Procedure

• Commercial silver nanoparticle ink is diluted with deionized water at a ratio of 3:1, and ultrasonic dispersion is employed to eliminate agglomeration.
• Plasma pretreatment of the alumina substrate enhances adhesion, while the workbench maintains a constant temperature of 90°C to enable real-time solvent removal.
• Ultrasonic atomization aerosol system, with a carrier gas flow of 22–28 SCCM and a sheath gas flow of 45–55 SCCM, using a 150 μm ceramic nozzle for layer-by-layer printing; after every 500 μm of deposition, the print head is raised to maintain the standoff distance.
• AutoCAD designs the raster path, controlling the number of print passes and adjusting the airflow to regulate the thickness of the truss and its overall density.
• The green body is printed and then sintered at a constant temperature of 350°C for 2 hours, during which the nanoparticles fuse to form a continuous metallic truss.
• SEM is used to observe microscale features, while a universal micro-compression testing machine performs quasi-static mechanical calibration, with an accompanying laser displacement sensor for acquiring deformation data.

(II) Five Core Advantages of Aerosol-Based Additive Manufacturing (Compared with Two-Photon Polymerization, Metal SLS, and Direct-Writing Printing)

• Support-free free-form shaping, eliminating the need for sacrificial formwork.

Unlike lithographic demolding and resin‑template metal‑plating approaches, this method leverages the surface tension of droplets to achieve self‑supporting structures, eliminating the need for subsequent etching to remove the substrate and significantly streamlining the fabrication process.

• Exclusive forming capability for near-solid micron-scale trusses

Two-photon lithography produces only hollow channels, whereas conventional inkjet printing can only fabricate coarse conductive traces; AJP enables the fabrication of near‑solid metal trusses with dimensions of 30–60 μm, offering superior current‑carrying and mechanical load‑bearing capabilities.

• Digital topology rapid switching, mechanics tailored on demand

By adjusting the printing path and sheath‑gas flow rate, one can fine-tune the cellular architecture, truss diameter, and overall density. On the same day, two types of deformable lattice structures—bending and buckling—are fabricated, maximizing iterative efficiency.

• Low-temperature nano-sintering, compatible with a wide range of metals and functional inks.

Silver, gold, conductive polymers, and ceramic inks can all be printed via atomization, enabling the fabrication of multifunctional micro‑arrays for energy storage, biotechnology, and catalysis. The sintering temperature is as low as 350°C, significantly lower than that required for metal‑based additive manufacturing.

• Microscale high precision, small-batch low-cost R&D

Micron‑scale, controllable cellular units enable the construction of experimental platforms on desktop‑level equipment, eliminating the need for large‑scale, dust‑free metal 3D‑printing production lines and significantly reducing the cost of research prototypes.

III. Five Key Technological Innovations

Innovation 1: A complete process for fabricating unsupported aerosol‑based 3D metal micro‑lattices.

   Innovation 2: A Digital Design System for Deformable Topologies

Innovative 3‑D Adaptive Ordered Lattice Correction Semi-Empirical Mechanical Model

Innovative four-stage pore‑controlled sintering mechanism

Innovative 5-Defect-Tolerant Microarray Fabrication Scheme

IV. Current Industry Challenges in the Process and Corresponding Breakthrough Solutions

Challenge 1: Conventional microarray fabrication requires sacrificing the template, resulting in complex processes and high costs.

Challenge 2: Existing micro‑trusses are predominantly hollow, resulting in poor mechanical load-bearing capacity and limited electrical conductivity.

Challenge 3: The periodic oscillations of the ordered lattice stress platform cannot be fitted by conventional foam models.

Challenge 4: Achieving different energy-absorption characteristics requires retooling and remanufacturing, resulting in a long iteration cycle.

Challenge 5: Metal‑melting 3D printing equipment is expensive, and achieving micron‑level precision remains difficult.

V. Future Diversified Industrial Application Scenarios

1 Three-dimensional porous electrode for micro lithium-ion batteries

2 Aerospace Lightweight Energy-Absorbing Metamaterials

3 Biomedical Bone Repair Scaffolds

4 MEMS micro-force sensing array

5 Microfluidic Catalytic Reaction Carriers

6 University-Level Metamaterials R&D Platform

Figure 1. AJ‑printed three-dimensional microlattice structures without support scaffolds. (A) Schematic illustration of the operating principle of the AJ 3D printer: ultrasonic actuation generates microdroplets containing metal nanoparticles; these droplets are carried by a carrier gas to the nozzle and focused onto the substrate surface by a sheath gas. (B) Schematic of the drop‑by‑drop, layer‑by‑layer printing process (left), along with a diagram of the force‑balance mechanism between cured and freshly deposited droplets. (C) Schematic of a typical printing path generated in AutoCAD (left) and a corresponding representation of the printed structure derived from that path (right). (D) Original two‑dimensional AutoCAD design (left) and an enlarged view illustrating the single‑layer printing sequence; after each layer is completed (indicated by the green arrow), the print head retracts inward (red arrow), and this cycle repeats until the entire layer is finished (green arrow), with the final three‑dimensional structure shown on the right. (E) A three‑dimensional AutoCAD model of a microlattice structure featuring octahedral unit cells. (F) Schematic of the compression‑testing apparatus for AJ‑printed microlattice structures. (For details regarding color coding in the figure legends, please refer to the online version of this article.)

Figure 2. CAD models and corresponding microlattice structures obtained via AJ printing. (A) Microlattice with hexagonal unit cells; (B) Microlattice with octahedral unit cells; (C) Microlattice featuring a central strut architecture and employing octahedral unit cells. By adjusting the sheath‑gas pressure, the thickness of the truss members in each lattice was modulated to achieve a specified volume fraction. For all three topologies investigated in this study, the diameter variation of truss members within a single microlattice was kept within ±8%.

Table 1 Physical properties of the 29 AJ‑printed microlattice samples used in this study. The table includes geometric parameters derived from scanning electron microscopy (SEM) images; detailed SEM images for each sample are provided in Supporting Information Figure S2. Relative density for each microlattice was calculated using mass measurements obtained with an analytical balance. Dimensional data in the table were employed to compute stress and strain values during compression testing. All samples were sintered in an oven at 350°C for 2 hours.

Figure 3. Microstructural analysis of sintered microlattice architectures. (A) SEM image of a 5 × 5 × 3 microlattice with hexagonal unit cells; (B) SEM image of a 4 × 4 × 4 microlattice featuring a hexagonal topology; (C) SEM image of a 4 × 4 × 4 microlattice with octahedral unit cells supported by vertical struts. The truss diameter and relative density (σ/σs) of these structures are both tunable. (D) Representative octahedral unit cells at different strut radii, along with their microstructures, compared to the printing path (red dashed line); (E) Schematic illustration and SEM image showing selectively etched outer regions of hollow struts using a focused ion beam. The scale bar in (D) corresponds to 50 µm. (For interpretation of the color codes in the figure legend, please refer to the online version of this article.)

Table 2 Comparison of Three-Dimensional Micro-Lattice Structures Fabricated by Different Manufacturing Processes

Figure 4. Comparison of the mechanical responses of differently AJ‑printed microlattice structures: (A) The initial shapes of two octahedral specimens; (B) Their stress–strain curves under uniaxial compressive loading; (C) The initial shapes of two octahedral specimens with central struts; (D) Their stress–strain curves under uniaxial compressive loading.

Figure 5. Relationship between the relative stiffness and compressive strength of AJ‑printed microlattice structures as functions of their relative density. Uniaxial compression test data show: (A) the variation of the relative stiffness of AJ‑printed microlattices with relative mass density; (B) the variation of the plateau stress with relative mass density. Linear scaling curves for tension‑dominated structures and quadratic scaling curves for bending‑dominated structures are also plotted for comparison.

Figure 6. Finite element modeling of an octahedral lattice. (A) Finite element model of the lattice with boundary constraints and applied loads (left) and a schematic of the compression test (right). (B) Schematic of finite element models of lattices composed of varying numbers of identical unit cells, ranging from 1 × 1 × 1 to 12 × 12 × 12. (C) Stress–strain curves obtained via finite element analysis for the lattices shown in (B). (D) Young’s modulus and plateau stress values corresponding to the data in (C).

Figure 7. Comparison of mechanical performance between finite-element analysis predictions and experimental observations. (A) For the octahedral microlattice structure under uniaxial compressive loading, comparison of the global stress–strain response predicted by the finite-element model with experimental data; (B) A frame-by-frame detailed comparison of finite-element analysis results and experimental data, illustrating the deformation, buckling, and collapse of local truss members and unit cells. Dashed boxes, ellipses, and circles highlight localized regions used for visual comparison; (C) Comparison of finite-element model predictions and experimental data for the octahedral microlattice structure with a central strut under uniaxial compression; (D) A frame-by-frame detailed visualization of the data in (C). Because the failure mechanisms in (C) and (D) are dominated by buckling—a form of instability—the finite-element analysis fails to accurately capture the pronounced oscillations observed in the stress‑plateau region; moreover, defects introduced during additive manufacturing lead to layer‑by‑layer failure of the microlattice, whereas the finite-element model predicts only a more uniform collapse process.

Figure 8. Validation of finite-element analysis results showing the dependence of mechanical response on relative density. (A) Relative elastic modulus of octahedral specimens; (B) Variation of relative strength with relative density, along with the corresponding finite-element model constructed using the same geometry as the experimental data.

Figure 9. Improved theoretical model of the stress–strain behavior of cellular materials. The theoretical model proposed in this study was compared with both the experimental data and the models developed by Avalle et al. [55] for O‑8(A), VP‑2(B), and HX‑5(C). The model accurately captures the periodic hardening and softening observed in the stress plateau region of the experiments. All models were obtained through manual curve fitting, with the quality of the fit evaluated using the r² statistic.