I. Abstract of the Paper’s Core Research

This study employs gold nanoparticle ink in conjunction with an AJP layer-by-layer stacking process to fabricate freestanding gold microcolumns with an aspect ratio of 10:1. Sintering temperatures were systematically varied across a range of 300–900°C, and the microstructural and compressive mechanical properties of two-dimensional gold films and three-dimensional microcolumns produced using the same process were compared. A biphasic micromechanical model was developed, and the key findings are as follows:

Aerosol without support 3D molding 10:1 High aspect ratio gold micropillars

Layer-by-layer deposition combined with in-situ pre-curing at 300°C eliminates the need for sacrificial supports; individual micropillars have a diameter of 100 μm and a height of 1 mm, making them well-suited to meet the three-dimensional probing requirements of neural probes.

Three-dimensional / The grain evolution in two-dimensional structures is entirely different.

Under identical sintering conditions, the average grain size of 3D micro‑columns is significantly smaller than that of two‑dimensional gold films; at 300°C, the grains measure only 26 nm, while at 900°C they grow to 1.7 μm, with porosity concurrently decreasing from 14.5% to 2%. The three‑dimensional, tightly packed stacking environment suppresses the free growth of grains.

Precise control of sintering temperature to tailor mechanical deformation modes.

Low-temperature sintering at 300–500°C: fine grains with high porosity, exhibiting brittle fracture upon instantaneous compression, and achieving a maximum yield strength of up to 52.4 MPa.

600°C transition zone: brittle–ductile mixed deformation;

At high temperatures of 700–900°C: coarse grains with low porosity undergo plastic buckling and can sustain large strains without instantaneous fracture.

Grain size is the dominant factor governing mechanical properties.

A grain–pore two-phase model was established, demonstrating that grain size exerts a stronger influence on compressive strength and elastic modulus than porosity. Fine grains impede dislocation motion, thereby enhancing strength, whereas coarse grains facilitate plastic deformation.

Biomechanical Compatibility Verification of Implants

The critical buckling force of the microcolumns is far greater than the 0.1–1 mN load required for brain tissue puncture, ensuring that they do not fracture during implantation and enabling the mass production of microelectrode arrays with varying stiffnesses.

II. Aerosols Complete Preparation Process for 3D Gold Micropillars

1) Commercial 2–5 nm gold nanoparticle ink, ultrasonically dispersed and compatible with ultrasonic atomization aerosol systems;

2) Plasma cleaning of the alumina substrate, followed by in-situ pre-curing at a constant 300°C on the printing platform to prevent microcolumn collapse;

3) The Optomec AJ300 system employs layer-by-layer additive manufacturing, with the print head automatically compensating for height adjustments after each layer. A total of 70 layers were built to fabricate a 1-mm‑diameter microcolumn.

4) Gradient sintering in a tube furnace, with a heating rate of 1°C/min, holding at the target temperature for 4 hours, and a cooling rate of 2°C/min.

5) FIB cross-sectional imaging was used to observe grain porosity, and a uniaxial compression testing machine was employed to perform mechanical load–displacement characterization.

III. Core Advantages of 3D Printing with Three- and Five-Dimensional Aerosols (Comparison with FIB and Lithography)

Self-supporting, high-aspect-ratio three-dimensional structure without support

No molds or sacrificial layers are required; millimeter-scale gold micropillars can be directly printed, whereas FIB can only fabricate micron-scale, short pillars at a cost that is tens of times higher.

Digital Custom-Size Array

CAD enables free adjustment of micropillar diameter, height, and inter‑pillar spacing, facilitating rapid iteration of both soft and rigid electrodes and significantly shortening the R&D cycle.

Gentle layer-by-layer forming, with no damage to the substrate.

Non-contact aerosol deposition enables processing of both alumina and flexible substrates, and is compatible with a wide range of biocompatible substrates.

Consumable costs have been significantly reduced.

Localized deposition of precious-metal gold ink; full‑wafer sputtering and ion milling result in substantial material waste, making small‑batch R&D more cost‑effective.

Programmable regulation of mechanical properties

Brittleness and ductility can be switched simply by adjusting the sintering temperature, without the need to change inks or equipment; a single production line can accommodate multiple device types.

IV. Five Core Technological Innovations of This Study

Innovation 1: Aerosol‑free, support‑free fabrication of high‑aspect‑ratio gold micropillars at a 10:1 ratio

Breaking the conventional limitations of low height-to-width ratios in metal additive manufacturing, in-situ pre‑curing addresses the issue of layer‑stacking collapse, enabling the direct, integrated fabrication of three-dimensional noble‑metal microarrays.

Innovation 2 confirms the differentiated growth patterns of 3D and 2D gold nanostructures.

Under identical thermal histories, a three-dimensional closed stacking environment suppresses grain growth, thereby challenging the conventional approach of directly applying empirical models of thin-film grain evolution to bulk materials.

Innovation: A Novel Theoretical Framework for Grain‑Dominant Micro‑Mechanics

By distinguishing the dual-weighted effects of porosity and grain size, we identify grain size as the key controlling variable governing compressive behavior and develop a predictive two-phase mechanical model.

Innovative Four-Temperature-Controllable Brittle–Plasticity Reversible Regulation Scheme

Gradient sintering enables the on-demand fabrication of two distinct mechanical behaviors—brittle puncture and plastic buffering—tailored to suit different implanted biological tissues.

Innovative Low-Cost Mass-Manufacturing Paradigm for Brain-Computer Microelectrodes

Eliminating reliance on costly cleanroom FIB and lithography equipment, an aerosol‑based desktop system enables the rapid fabrication of high‑density neural probe arrays.

V. Current Industry Challenges and Corresponding Solutions for Overcoming Them

Challenge 1: FIB and lithography can only fabricate two-dimensional gold structures with low aspect ratios, making it impossible to perform three-dimensional probing of brain tissue.

✅ Aerosol-based layer-by-layer support-free printing, with 10:1 self-supporting microcolumns, enables multi-layer neural signal recording within the brain.

Challenge 2: Traditional metallic microstructures are mechanically limited—hard pillars readily puncture soft tissues, while soft pillars are prone to buckling and failure.

✅ Gradient sintering at 300–900°C enables precise control of grain size, enabling the on-demand fabrication of two types of micropillars: brittle‑piercing and ductile‑buffering.

Challenge 3: Confusing the evolution laws of three-dimensional and two-dimensional metallic microstructures, leading to indiscriminate process design.

✅The system compares the differences between thin films and microcolumnar grains, providing a dedicated sintering design basis for three-dimensional microstructures.

Challenge 4: Precious-metal microfabrication equipment is expensive, and small-batch custom production entails extremely high costs.

✅ Maskless direct writing of aerosols eliminates the need for vacuum and ion milling production lines, significantly reducing prototyping costs.

Challenge 5: Implantation of micro‑columns is prone to buckling and fracture, resulting in poor clinical suitability.

✅ Optimization: The micro‑pillar sintered at 400°C exhibits a yield strength of 52.4 MPa, significantly exceeding the puncture load, ensuring stable implantation without fracture.

VI. Future Diversified Industrial Application Scenarios

1 High-Density Brain-Computer Interface Neural Microelectrode Array

Brittle micro‑columns penetrate the dura mater to acquire deep neural signals, while pliable micro‑columns are designed for long-term implantation in soft brain tissue.

2 Miniature Electrochemical Biosensor

High‑surface‑area three‑dimensional gold pillars enhance biomolecule adsorption, enabling ultra‑sensitive trace‑level detection on a microchip.

3 Microfluidic Catalytic Reaction Array

Controllable-pore gold micropillars, used as miniature electrodes for small-molecule catalysis and electrochemical energy storage.

4 Implantable Flexible Medical Sensing Elements

Biodegradable substrate‑based composite microstructures for minimally invasive, long-term in vivo physiological signal monitoring.

5 University-Based R&D Platform for Precious Metal Micro- and Nano-Manufacturing

Rapid fabrication of three-dimensional metallic microstructures with diverse mechanical properties, for applications in biomechanics and the study of microcatalytic mechanisms.

Image and Text Guide

Figure 1. Schematic of the entire process for AJP‑fabricated 3D gold microcolumns: (a) the printing system, (b) a close-up of the deposition head, (c/d) SEM images of the microcolumns and FIB cross‑sections, (e) uniaxial compression testing.

Figure 1(a) shows ultrasonic atomization of gold nanoparticle ink, with N₂ carrier gas delivered to the deposition head; the X–Y stage moves while the deposition head advances layer by layer in the z‑direction. Figure 1(b) is a close-up of the deposition head: the sheath gas (N₂) focuses the ink mist outside a 150 μm nozzle, achieving single‑droplet resolution of 10–15 μm. Figure 1(c) presents an SEM image of a 1 mm‑high gold microcolumn after printing, and Figure 1(d) shows a FIB‑cut cross‑section revealing an outer diameter of ~100 μm and a central hollow cavity of ~40 μm. Figure 1(e) illustrates a uniaxial compression apparatus: a load cell at the base supports the column’s bottom, while an upper indenter applies downward force, with a high‑resolution camera recording the process. At low sintering temperatures, brittle fracture occurs; at higher sintering temperatures, ductile fracture is observed, corresponding to microstructures characterized by fine grains and high porosity versus coarse grains and low porosity.

Figure 2. Weibull distribution of grain sizes in AJ-printed 2D gold thin films sintered at 300/600/900 °C for 4 h.

All three histograms were fitted using a two-parameter Weibull distribution (red lines). At 300 °C: μ = 298 nm, λ = 332.33, k = 3.39, N = 51; at 600 °C: μ = 424 nm, λ = 469.61, k = 3.58, N = 41; at 900 °C: μ = 3314 nm (3.3 μm), λ = 3565.51, k = 5.99, N = 30. As the sintering temperature increases, the average grain size grows from 0.3 μm to 3.3 μm, and the width of the grain-size distribution expands markedly; the value of k > 3 indicates a left-skewed distribution.

Figure 3. SEM images of 3D gold micropillar FIB cross-sections: (a) after sintering at 300 °C, (b) at 600 °C, and (c) at 900 °C.

Each row shows: (i) the original SEM image, (ii) pores highlighted in red using ImageJ, and (iii) grain boundaries superimposed in yellow. At 300 °C, grains are fine (~26 nm), with a porosity of 14.9% and small, uniform pores; at 600 °C, grains grow to 193 nm and twin boundaries appear, while porosity decreases to 6.12% and pores become irregular; at 900 °C, grains reach 1.7 μm (with a maximum of 4.4 μm), porosity drops to just 3.11%, and only a few large, irregular pores remain. Figure 3(d) illustrates the three stages of sintering: Stage 1—Necking (formation of particle necks); Stage 2—Grain growth and defects; Stage 3—Large grains (complete densification with well-defined grain boundaries).

Figure 4. Quantitative statistics of 3D gold microcolumn grain size and porosity as functions of sintering temperature.

Figure 4(a): Superimposed histograms of grain sizes at three temperatures (with the x-axis on a logarithmic scale): at 300 °C, grains are concentrated between 10 and 80 nm; at 600 °C, the distribution broadens to 10 nm–1.5 μm; and at 900 °C, grains are predominantly in the 0.5–3 μm range. Figure 4(b): Three Weibull fits: μ = 25.56 nm / 192.80 nm / 1699.01 nm, N = 3194 / 788 / 104. Figure 4(c): Scatter plot with box-and-whisker plots: the grain size distribution at 600 °C exhibits the widest spread, showing a bimodal pattern. Figure 4(d): Porosity decreases nonlinearly with sintering temperature: 14.9% at 300 °C → 8.7% at 400 °C → 6% at 500–600 °C → 2.5% at 900 °C.

Table 2. Weibull grain-size parameters for the 3D gold microcolumns at all seven sintering temperatures

It covers seven temperature levels between 300 and 900 °C. The average grain size μ is: 26 nm at 300 °C → 46 nm at 400 °C → 179 nm at 500 °C → 193 nm at 600 °C → 355 nm at 700 °C → 821 nm at 800 °C → 1,699 nm at 900 °C. The scale parameter λ/μ remains essentially constant (1.13–1.14), indicating that the grain-size distribution shape does not change with temperature; rather, it undergoes a uniform shift. The shape parameter k is 2.00 at 300 °C (right-skewed), while at other temperatures it ranges from 1.14 to 2.33.

Figure 5. Force–displacement (stress–strain) curves from microcolumn compression, along with synchronized video recordings: (a) brittle fracture at 300 °C, (b) brittle–ductile transition at 600 °C, (c) ductile fracture at 900 °C.

Each column of four video screenshots corresponds to four points on the curve, labeled ①→②→③→④. At 300 °C: the material exhibits linear elasticity up to 0.25 N / 31 MPa, after which it fractures abruptly; the column breaks at its midsection (indicated by the failure arrow), with virtually no plastic deformation. At 600 °C: following the elastic regime, a period of plastic deformation is observed before sudden instability occurs. At 900 °C: after the elastic region, the stress decreases continuously with increasing displacement, and the column bends into an S‑shaped configuration like a soft metal, demonstrating the characteristic ductile buckling of gold.

Figure 6. Summary of the compressive mechanical behavior across all seven sintering temperatures: (a) family of stress–strain curves, (b) peak force and Euler buckling modulus, (c) initial slope.

Figure 6(a): The curve exhibits a sharp transition to brittleness between 300 and 500 °C; a plastic plateau emerges starting at 600 °C; and full plasticity is observed between 700 and 900 °C. Figure 6(b): The peak force reaches 345.8 mN at 400 °C, then decreases monotonically with increasing temperature to approximately 80 mN; the elastic modulus E_b calculated using Euler’s buckling theory also attains its maximum at 400 °C (3.59 GPa). Figure 6(c): The initial loading slope (apparent elastic modulus) likewise peaks at 400 °C and remains on the order of 300 MPa from 400 to 900 °C, following an inverse Hall–Petch relationship—larger grains correspond to lower modulus.

Table 3. Peak force, peak stress, elastic‑segment slope, and Young’s modulus of micropillars at different sintering temperatures

300 °C: peak force 205.2 ± 29.2 mN, peak stress 31.1 ± 4.4 MPa, slope 0.56, E_b = 2.13 GPa; 400 °C: peak force 345.8 ± 37.1 mN (the highest value across the entire table), peak stress 52.4 ± 5.6 MPa, E_b = 3.59 GPa; at 500 °C, the peak force drops below 143 mN; from 600 to 900 °C, the peak force remains stable at 70–95 mN, with peak stress ranging from 8.8 to 12.4 MPa. 400 °C marks a turning point in the material’s mechanical properties.

Figure 7. Two-phase (grain + porosity) mechanical model: (a) Average grain/porosity size vs. temperature, (b) Porosity-to-grain ratio R vs. temperature, (c) Relative modulus vs. porosity.

Figure 7(a): The grain size increases monotonically with temperature, while the average pore diameter reaches a peak of 33 nm at 400 °C and then gradually decreases to 24 nm as sintering progresses. Figure 7(b): The ratio R = d_P/d_G decreases monotonically from 0.63 at 300 °C to 0.025 at 900 °C, indicating that pores become progressively smaller relative to the grains. Figure 7(c): According to the two-phase model, the effective elastic modulus f_c = E_eff/E_0 declines with increasing porosity P; the blue data points represent experimental measurements. At low values of R (corresponding to higher sintering temperatures), the curve exhibits a more gradual decline, suggesting that even with reduced porosity, the dominant influence of grain size—resulting in fewer grain boundaries—leads to a more pronounced reduction in modulus.