I. Abstract of the Paper’s Core Research

This study develops a long‑lasting PDMS aerosol ink compatible with ultrasonic atomization, establishes an in‑situ thermal curing 3D‑AJP fabrication platform, and systematically investigates how substrate temperature, airflow focusing ratio, and printing offset influence the formation of two‑dimensional lines and three‑dimensional self‑supporting microstructures. Furthermore, it validates multi‑functional devices driven by mechanical, fluidic, and magnetic actuation. The key findings are as follows:

Long-term storage stability PDMS-specific aerosol ink

Using mesitylene as the diluent solvent, crosslinking of the A/B components is delayed, allowing the ink to remain stably atomizable after standing for 21 days. With viscosity maintained at 1.85 cP, it is well-suited for ultrasonic atomization, addressing the common issues of traditional PD inks—gelation and rapid degradation within a single day.

Support-Free Freeform Fabrication Process System

By leveraging a substrate with controllable high temperatures, droplet deposition enables instantaneous solvent evaporation and in-situ hydrosilylation crosslinking, eliminating the need for any sacrificial support layers. Under optimal conditions at 250°C, it can print self-supporting trusses with a minimum tilt angle of 36°, individual micro‑columns up to 2.2 mm tall with an aspect ratio of 22, and feature line resolutions as fine as 27 μm.

Temperature determines the molding design window.

At a substrate temperature of 75°C, three-dimensional structures cannot be assembled; within the 100–275°C range, as the temperature increases, the critical height of the micropillars rises, while the minimum self-supporting tilt angle decreases. COMSOL simulations reveal that the axial temperature gradient in the micropillars is the primary driver of tip instability.

A universal printing paradigm for polymeric materials

The same aerosol‑based equipment and molding logic can directly fabricate three‑dimensional micro‑columns made of Ecoflex, conductive PEDOT:PSS, polyimide, and SU‑8 photoresist, without requiring substantial adjustments to the ink formulation or process parameters, demonstrating exceptional material versatility.

Complete verification of three types of functional devices

① BCC‑structured 3D PDMS micropillar array: subjected to 10,000 cycles of compressive strain at 30% and 50%, it exhibits stable hysteresis loss and excellent fatigue resistance; ② Hollow, self‑standing microfluidic channels: operate at flow rates of 0.25–5.5 mL/min with no leakage and can withstand repeated bending; ③ SPION‑based magnetic microcolumns: demonstrate a stable, reversible deflection of 22.98° under an external magnetic field, suitable for use as soft artificial cilia and miniature actuators.

II. PDMS Complete Process Flow for 3D Aerosol Jetting

Ink preparation: Mix the PDMS base with a curing agent at a 10:1 ratio, dilute with mesitylene, and stir uniformly for 30 minutes to obtain a low-viscosity, atomizable colloid; directly incorporate superparamagnetic Fe₃O₄ nanoparticles into the magnetic composite ink.

Plasma activation of silicon wafer substrates, with an external PID‑controlled heating platform; temperature range adjustable from 75 to 275°C.

Optomec AJ300 aerosol system, 150 μm ceramic nozzle, nitrogen carrier and sheath gas co‑axially focused, ultrasonic atomization generates micron‑scale PDMS droplets;

Custom print paths in CAD allow you to control inclined trusses, lattice structures, and hollow-channel geometries via interlayer offsets.

Droplets undergo instantaneous crosslinking and curing upon contact with a high-temperature substrate, enabling layer-by-layer stacking of unsupported three-dimensional structures that can be directly used for device testing after room-temperature cooling.

Compression fatigue, fluid transport, and magnetic-field-driven performance were characterized separately.

III. Core Advantages of the Three‑Five Aerosol Process (Comparison with Soft Lithography, DIW, and Two‑Photon Lithography)

No loss of support throughout the entire process, eliminating the need for demolding and a full suite of post-processing steps.

Unlike FRESH suspended‑droplet printing and two‑photon‑induced direct‑write techniques, this method employs single‑layer, layer‑by‑layer thermal curing to directly fabricate suspended, inclined microstructures, thereby eliminating the risk of micro‑pillar fracture or damage during demolding.

Ultra-wide material compatibility, conductive / elastic / General-purpose photosensitive polymer

A single system is compatible with insulating elastomers, conductive polymers, high‑temperature‑resistant polyimides, and UV‑curable photoresists, enabling the rapid fabrication of multifunctional composite microdevices.

Micrometer-level ultra-high resolution + Ultra-high aspect ratio

Minimum feature size: 27 μm; aspect ratio of self-supporting micro‑columns: 22—far exceeding that of direct‑write DIW (up to 10)—balancing fine patterning with substantial height.

Digital rapid iteration of freeform three-dimensional structures

No mold opening is required; with CAD, you can modify dot patterns, hollow channels, and magnetic micro‑pillar arrays with a single click, enabling the production of multiple structural comparison samples in just a few hours and significantly boosting R&D efficiency.

Gentle forming is compatible with flexible substrates, resulting in lower equipment costs.

No expensive UV laser source is required; deposition occurs at ambient pressure and without contact, and silicon, metals, and flexible thin films can all serve as printing substrates, resulting in a low barrier to laboratory implementation.

IV. Five Core Technological Innovations of This Study

Innovation 1: Long-lasting, stable oil-based ink specially formulated for PDMS aerosols containing mesitylene.

A solvent‑based crosslinking‑retarding formulation addresses the common issue of PDMS two‑component inks gelling and clogging nozzles, extending the shelf life to 21 days and enabling long‑term, continuous mass‑production printing.

Innovation 2: In-situ Thermosetting Support-Free 3D-AJP Forming Mechanism

The mechanism of rapid droplet evaporation and crosslinking–induced shaping on high-temperature substrates is elucidated, and a quantitative process window—spanning temperature, critical pillar height, and minimum self-supporting angle—is established, providing a theoretical foundation for the fabrication of three-dimensional elastic structures.

Innovation 3: A standardized process for universal polymer-based aerosol printing

The unified atomization and thermal curing process is compatible with a wide range of materials, including silicone rubber, conductive polymers, and high‑temperature‑resistant polyimide, thereby overcoming the limitations of single-material printing.

Innovative Integrated Multifunctional Microdevice Manufacturing Solution

A single fabrication process simultaneously produces three types of devices—fatigue‑resistant elastic lattice structures, leak‑free microfluidic channels, and magnetically responsive soft actuators—spanning the two major fields of microfluidics and soft robotics.

Innovative Simulation Model of Temperature-Gradient-Induced Instability

By integrating COMSOL finite-element simulations, we elucidate the mechanism by which temperature reduction at the micro‑pillar tip leads to divergent collapse, enabling accurate prediction of the ultimate forming height and reducing experimental trial-and-error.

IV. Current Industry Challenges and Corresponding Solutions for Overcoming Them

Challenge 1: Conventional PDMS printing inks tend to gel prematurely, resulting in an extremely short print window.

✅ The mesitylene‑diluted system inhibits hydrosilylation, allowing the ink to remain stably atomizable after three weeks of storage and significantly reducing consumable waste.

Challenge 2: Three-dimensional elastic microstructure fabrication requires support layers, and demolding can easily damage delicate features.

✅ The substrate undergoes instant thermal curing, with each layer solidifying immediately upon deposition. A 36°‑inclined gantry enables stable part formation without auxiliary support, and no post‑processing damage occurs.

Challenge 3: DIW suffers from poor direct-write resolution and low aspect ratio, while two-photon lithography systems are prohibitively expensive.

✅ Aerosol‑based beam focusing enables 27‑μm fine lines with an aspect ratio of 22, while the equipment cost is significantly lower than that of two‑photon polymerization systems.

Challenge 4: A single forming process is compatible with only one type of polymer, while composite devices require multi‑equipment processing.

✅ By simply switching inks on the same printing platform, conductive, elastic, and high‑temperature‑resistant microstructures can be fabricated, enabling integrated heterogeneous fabrication.

Challenge 5: There is no quantitative theory for the maximum forming height of three-dimensional micro‑columns; it relies entirely on empirical trial and error.

✅ A substrate‑temperature–microcolumn‑temperature gradient simulation model has been established, enabling precise prediction of the critical upright height and allowing standardized process windows to be directly reused.

VI. Future Diversified Industrial Application Scenarios

1 Organ-on-a-chip, microfluidic biochips

Self‑standing hollow PDMS microchannels and multilayer three‑dimensional fluidic networks for drug screening and organoid culture in microfluidic systems.

2 Soft-Robot Magnetic Micro-Actuators

SPION composite magnetic microcolumn arrays, with remotely controllable bending under an external magnetic field, serve as core components for artificial cilia and miniature grasping robots.

3 Flexible Tactile Sensor Array

High-fatigue‑resistant three-dimensional PDMS micro‑lattice, suitable for wearable pressure sensing and as a cushioning sensing layer in biomimetic skin.

4 Miniature Flexible Electrochemical Devices

PEDOT:PSS conductive three-dimensional electrodes, flexible micro-supercapacitors, and biosensing microelectrode arrays.

5 General R&D Platform for Micro- and Nanomanufacturing in Universities

Low-cost aerosol-based equipment enables rapid fabrication of diverse polymeric three-dimensional microstructures, serving as an alternative to expensive two-photon and soft-lithography production lines.

Image and Text Guide

Figure 1. Overall process for AJP printing of 3D PDMS microstructures

Figure 1(A) shows a high‑frequency ultrasonic atomizer that nebulizes PDMS ink diluted in mesitylene into micron‑scale droplets; Figure 1(B) depicts droplets approximately 3–5 μm in size, with nitrogen carrier gas fed into the print head; Figure 1(C) illustrates a coaxial sheath gas that focuses the aerosol into a narrow beam, which is ejected through a 150‑μm nozzle; Figure 1(D) shows droplets landing on a heated substrate, where solvent evaporation and in‑situ crosslinking occur instantaneously, enabling layer‑by‑layer deposition to form a 3D BCC micro‑lattice—each strut has a diameter of about 100 μm, without the need for any sacrificial supports.

Figure 2. PDMS ink formulation, linewidth optimization, and 2D complex patterns

Figure 2(A): Ink formulation process: 2 g PDMS base + 18 mL mesitylene, vortexed for 10 min; hardener added at a 10:1 ratio; followed by roll‑up in a vial for 30 min. Figure 2(B): Line width versus focus ratio (FR): three curves at 150/200/250 °C; as FR increases from 1 to 8, line width decreases from ~400 μm to ~27 μm. At 250 °C with FR = 8, the best resolution is 27 μm; for practical applications, FR = 2 is preferred. Figure 2(C): 2D demonstration: CMU logo and Scotty Dog outline (10 layers stacked, feature size ≤100 μm). Figure 2(D): Andrew Carnegie facial relief (10 layers, with 500 μm‑scale facial details).

Figure 3. Formation mechanism of the three-zone PDMS micropillar array and its temperature window.

Figure 3(A): Three‑zone microcolumn architecture: I, the substrate‑formation zone (impact + overspray, thermal shock); II, the stable vertical‑growth zone (cures on the fly, with each droplet solidifying within 1–2 s); III, the spreading zone (insufficient tip temperature, causing droplets to flatten into a bulb). Figure 3(B): 2×2 microcolumn arrays at substrate temperatures of 75–275 °C: at 75 °C, no columns form; at 100–200 °C, straight columns develop; at 275 °C, columns grow up to 2200 μm in height, with an aspect ratio of ~22. Figure 3(C): COMSOL simulation of straight‑column height h′ as a function of cross‑sectional temperature: h′ increases exponentially with temperature, with an optimal tip‑temperature window of 31 °C (for a 150 °C substrate, from 97 °C to 66 °C; for a 225 °C substrate, from 97 °C to 66 °C). Figure 3(D): After 21 days of storage under identical conditions, column height and diameter remain nearly unchanged (pot life ≈ 3 weeks).

Figure 4. Design space of self-supporting inclined struts, hollow microchannels, and material universality

Figure 4(A): Inclined struts are realized by applying incremental lateral offsets δ in CAD; at 150/200/250 °C, the critical self-supporting angles θc are 60°/46°/36°, with δ = 4/5/6 μm—higher temperatures enable printing of flatter inclined beams. Figure 4(B): Hollow PDMS microchannels: in CAD, the centerline circle diameter d ranges from 40 to 200 μm, while the measured outer diameter D increases linearly with d; higher temperatures yield a larger D/d ratio (with a maximum deviation from 1:1 at 250 °C). Figure 4(C): Material substitution within the same process: polyimide 5×5 lattice, PEDOT:PSS micropillar array, Ecoflex micropillars (aspect ratio 15), and SU‑8 micropillars (aspect ratio 7)—only the ink’s solvent/resin is changed, while the process framework remains universal.

Figure 5. Mechanical properties of a 5×5 BCC PDMS micropattern

Figure 5(A): A 5×5 BCC micro‑lattice printed at 250 °C; unit cell dimensions are 2.3 × 2.3 × 1.1 mm, with an average strut diameter of 87 ± 3 μm; converging struts meet at the nodes, while diverging struts radiate outward from the nodes. Figure 5(B‑i/ii): Compression stress–strain curves after 5,000 cycles: at 30% strain, the hysteresis energy density rapidly decreases from 13.76 × 10⁴ to ~12 × 10⁴ J/m³ (Mullins effect); at 50% strain, it stabilizes at 25.6 × 10⁴ J/m³. Figure 5(B‑iii): Optical deformation: at 30% strain, the second‑layer struts exhibit early buckling; at 50% strain, the upper‑layer struts collapse into a nearly solid state. Figure 5(B‑iv): SEM images after 10,000 cycles: the structure remains intact with no cracks, demonstrating excellent fatigue resistance.

Figure 6(A)(i): Microfluidic experimental setup: syringe pump with a flow rate of Q = 0.25–5.5 mL/min; pressure drop estimated using the Hagen–Poiseuille equation, ΔP = 8μLQ/πr⁴; (ii–iv): Schematic illustration of leakage detection on a stainless-steel substrate. Figure 6(B)(i): Physical device; (ii): Optical image of droplets at the channel apex. Figure 6(C)(i–vi): Time-lapse sequence showing stepwise injection of cresol red dye into a 3D hollow microchannel, demonstrating leak-proof performance. Figure 7(A): Optical image of SPION‑functionalized magnetic PDMS micropillars, along with an ink bottle.

Figure 7(C): Magnetic field B of an NdFeB magnet versus the deflection angle θ of the micropillars: at 450 mT, the deflection is 23°; on the right, a 2×2 array undergoes gradual deflection as the magnetic field increases from 0 to 450 mT. Figure 7(D): Hysteresis loop of SPIONs (soft magnetic material). Figure 7(E): θ versus body force, along with the FEA‑predicted Von Mises stress distribution (corresponding to a deflection of 22.98°).