I. Core Research Summary

• Integrated Ultra-Thin Multilayer Sensing Structure for Aerosols

The entire sensor stack has a total thickness of only 12.4 μm, fabricated via aerosol jet printing in three layers—upper and lower silver electrodes sandwiching a PI dielectric. By optimizing the PI ink ratio to 4:1, a dense 2.7‑μm insulating layer can be formed in just three printing passes, with no interlayer short circuits even after multiple layers are printed. The electrode resistance is below 25 Ω, and the device exhibits excellent high‑frequency sensing performance.

• Ultra-high stretchable mechanical adaptability

A serpentine conductive pattern design, validated through finite-element analysis and mechanical testing, can withstand 250% elongation and 180° bending. After compression and release within the implantation catheter, it exhibits no open‑circuit failures and conforms to intracranial narrow vessels measuring 3–5 mm, without disrupting native hemodynamics.

• Passive, battery-free, long-range wireless sensing

The sensor employs inductive coupling resonance with an external coil to acquire signals, eliminating the need for on-chip energy storage. It achieves a readout range of 8 cm in air, 7 cm in saline, and up to 6 cm in a muscle‑tissue‑simulating environment—significantly surpassing conventional implantable sensors (≤3 cm)—while keeping electromagnetic emissions well below medical safety limits.

• Precise blood-flow–responsive sensing mechanism

The impact of blood flow within the aneurysmal sac induces deformation of the PI dielectric layer, leading to a synchronous increase in capacitance and a linear shift in the resonant frequency. For a 5-mm vessel, the sensitivity is 0.29 pF·s/m, while for a 3.5-mm microvessel it is 0.084 pF·s/m, enabling precise discrimination of physiological blood flow velocities in the range of 0.05–1 m/s.

• All-additive low-cost manufacturing

By eliminating highly complex microfabrication processes such as photolithography, evaporation deposition, and bonding, CAD‑based digital tuning enables precise adjustment of sensor dimensions. A single aerosol‑based process achieves full‑layer patterning of both conductive and insulating materials, allowing for the mass production of miniature sensors tailored to various vascular specifications.

II. Core Operating Principle of Aerosol Spraying and Its Unique Process Advantages

(1) Complete 3D multi-layer sensor printing process

• PMMA sacrificial layer is spin-coated onto a glass substrate, with the platform preheated to 80°C; pneumatic atomization is used to deposit PI insulating ink, and a 300-μm nozzle prints the underlying support layer, which is then cured at 240°C.
• Ultrasonic atomization of silver nano‑ink, 200 μm nozzle printing of a serpentine conductive electrode layer, followed by high‑temperature sintering at 240 °C to reduce electrical resistance;
• Aerosol printing is used again to deposit an ultra-thin insulating layer in the middle of the PI, isolating the upper and lower silver electrodes to prevent leakage current.
• Top-layer silver electrodes are formed concurrently with aerosol deposition, and secondary curing completes the full capacitor structure.
• Acetone dissolves the PMMA to detach the sensing film, which is then transferred onto an Ecoflex flexible elastomer; after encapsulation, it is mounted onto an interventional stent.

(II) Five Core Process Advantages (Compared with Lithography / Vapor Deposition / Film Transfer)

• Integrated molding of conductive and insulating materials in a single piece of equipment.

Pneumatic atomization PI and ultrasonic atomization of silver nanoparticles are compatible in a dual-mode configuration, enabling multi-layer stacking of dielectrics and electrodes on a single platform—eliminating the need for multiple vacuum and coating systems and significantly reducing capital investment in the production line.

• Ultra-thin, uniform microstructures fabricated by direct writing, with mask-free, free-form design.

Micrometer-scale serpentine electrodes and ultra-thin insulating layers are fabricated via digital direct writing, enabling on-demand tuning of the sensing area and line dimensions to meet the miniaturization requirements of 3–5 mm intracranial microvessels.

• Non-contact, gentle deposition, compatible with flexible transfer substrates.

The nozzle maintains a stable standoff distance from the substrate, avoiding mechanical compression that could damage the thin sacrificial layer; after printing, it can be cleanly peeled off to expose the medical-grade elastomer without compromising the conductive sensing pathways.

• Multi-layer precision isolation addresses the pain point of electrode short circuits.

The sheath airflow precisely controls the width of the mist stream, enabling layered and segmented curing of the PI insulating film. Even at a thickness as low as 2.5 μm, it can fully cover the protrusions of the silver electrodes, thereby preventing electrical shorting between the upper and lower metal layers.

• Low-loss nano‑ink deposition reduces the cost of biomedical implants.

On-demand regional additive manufacturing eliminates the need for full‑area coating, with precious‑metal silver ink consumption far lower than that of spin‑coating or sputtering, enabling rapid prototyping and iterative production of small‑batch, custom medical samples.

III. Five Key Technological Innovations

Innovation 1: A dual‑atomization, partitioned aerosol multi‑layer printing system

Innovative 2 4:1 aerosol ink formulation specially designed for polyimide

Innovative 3-in-1 integrated flexible implantable device with sacrificial layer transfer

Innovation 4: Inductively Coupled Long-Range Passive Sensing with Matched Aerosol Low-Resistance Electrodes

Innovative 5‑Snake‑Shaped Highly Stretchable Aerosol‑Based Conductive Pattern Design

IV. Existing Process Challenges and Corresponding Solutions for Overcoming Them

Challenge 1: During multilayer printing, the silver electrodes protrude, and the thin PI dielectric cannot fully cover them, leading to interlayer short circuits.

Challenge 2: High-temperature sintering can easily damage the PMMA sacrificial layer, making it impossible to cleanly delaminate the sensor.

Challenge 3: Conventional aerosol‑printed conductive lines exhibit high brittleness, leading to the rupture of the conductive pathways upon stretching.

Challenge 4: Ink atomization is unstable, leading to inconsistent line widths during prolonged printing and poor sensor consistency.

Challenge 5: Implanted metals are prone to biocorrosion and signal drift.

V. Future Diversified Industrial Application Scenarios

• Long-term post-interventional monitoring of intracranial aneurysms

• Miniature Sensor-Stented Vascular Intervention Device for the Entire Body

• Flexible Implantable Organ Pressure Sensing

• In vivo neurovascular research model in animals

• Wearable subcutaneous passive health-monitoring patch

Figure 1. AJP fabrication and material characterization. A) Schematic and micrograph of PI deposition via pneumatic atomization to form an AJP structure (left); magnified view showing the surface morphology of the printed PI (right). B) Overall process of silver deposition using ultrasonic atomization (left) and a magnified image of the silver patterned structure on a PI substrate (right). C) Schematic illustrating the fabrication of an implantable flow sensor in an aneurysm model using AJP technology; inset shows the multilayer architecture of the sensor package. D) SEM cross-sectional image revealing the sensor’s multilayer structure. E) SEM images of silver nanoparticles before printing (left) and after sintering (right), highlighting the distribution of agglomerates. F) X-ray diffraction analysis of sintered silver nanoparticles on a glass slide, as well as the bottom and top silver electrodes on a PI substrate: no changes in crystalline structure were observed for any of the three samples.

Figure 2. Optimization of the AJP process for thin, multilayer, stretchable biosensors. A) Cross-sectional view of the printed PI layer, showing that a 4:1 ink ratio is optimal for achieving thicker layers without increasing feature width; B) Microscopic images after one and five prints using a 1:1 PI ink ratio, demonstrating reduced feature width; C) Microscopic images after one to five prints with a 4:1 PI ink ratio, clearly revealing the feature structure; D) Schematic illustrating the transfer of printed material onto silicone rubber; E) Resistance comparisons before and after transfer, indicating negligible resistance changes; F) Repeated printing of silver nanoparticles (AgNPs) reduces resistance values; G, H) Lowering the stage temperature from 80°C to 50°C enhances the roughness of the printed AgNP layer; I) When the stage temperature is reduced from 80°C to 50°C, the resistance of the sintered AgNPs remains unaffected; J) Cross-sectional views of the dielectric layer deposited on both the underlying PI and the AgNP layer reveal a dielectric thickness of approximately 2.5 µm; K) The cross-section of the finished sensor exhibits a low-profile morphology, with a maximum thickness of 12.4 µm.

Table 1. Printing and Material Parameters

Table 2. Sensor Dimensions Composed of a Multilayer Structure

Figure 3. Mechanical modeling, fabrication, and reliability analysis of a printed biosensor for intravascular catheter implantation. A) A highly ductile sensor was designed via finite-element modeling and compared with an experimentally fabricated counterpart; B, C) Measurements show that the printed sensor exhibits negligible changes in electrical resistance under cyclic bending and radial tensile loading; D) Leveraging its highly flexible and ductile architecture, the sensor can be deployed during catheterization within the target vessel.

Figure 4. Parameter study and optimization of the inductive coupling approach. A) Schematic of a battery-free wireless system, comprising an implantable flow sensor and two external antenna coils; B) Schematic of the device architecture and an example excitation signal; the received transient sensor signal differs between resonant on- and off‑states; C) Experimental validation of the optimal number of turns in the sensor coil required to maximize signal amplitude at the resonance frequency; D) Maximum resonant signal amplitude as a function of frequency (blue curve), corresponding to a 9‑turn excitation coil and a 5‑turn receiving coil; E) Minimization of noise by tuning the antenna coil parameters within the target frequency range; F) Signal amplitude indicating that sensors with thicker silver layers (multi‑layer plating) yield larger resonance peaks; G) Comparison of the excitation magnetic field strength at a distance of 1 cm from the excitation coil (averaged over a 20 × 20 cm² projection area; safety limit: 1.6 A·m⁻¹); H) The CardioMEMS high‑frequency system exhibits a higher ERP value than the 5 V used in this study; I) In vitro test results for the sensor coil and ceramic capacitor under various conditions, including frequency sweeps of the sensor coil in air and readout distances up to 8 cm; J) Readout distance of up to 7 cm when the sensor coil is immersed in saline solution; K) Detection range of up to 6 cm within human tissue.

Figure 5. In vitro evaluation of the printed sensor’s performance in a biomimetic vascular model. A–C) Photographs of sensors integrated into a medical stent within an aneurysm model: A) Overall magnified view of the sensor; B) Oblique magnified view of the ultra-thin, low-profile sensor; C) Copper‑coil‑connected sensor for wireless monitoring of hemodynamics in the biomimetic model; D) Pulsatile flow‑rate waveform captured by the sensor: the average capacitance increases as the flow rate rises from 0.05 m·s⁻¹ (left) to 0.35 m·s⁻¹ (right); the schematic on the right illustrates the mechanism by which increased blood flow within the aneurysm induces changes in capacitance; E, F) Trends of capacitance versus increasing flow rate in a 5‑mm vessel (E) and at higher flow rates in a 3.5‑mm vessel (F); G) Resonant frequency decreases with increasing flow rate; H) Continuous frequency‑sweep plot showing the shift in resonant frequency; I) Resonance peak amplitude diminishes with increasing detection distance (a resonance peak is observable only at detection distances of 6 cm or greater); J) When monitoring resonant frequency with a VNA, the required readout distance is markedly reduced (to less than 1 cm), demonstrating the superiority of the optimized wireless inductive coupling approach.

Table 3. Comparison of Parameters for Inductively Coupled Wireless Sensor Devices

a) The ratio of the wireless read range to the sensor’s cross-sectional area was defined. Our results indicate that, among the reported inductive coupling methods, this approach exhibits the best performance; b) It should be noted that some literature provides limited information on sensor dimensions. Units are as follows: length (l), diameter (d), width (w), and thickness (t); c) R.f.: resonant frequency; d) High-frequency signals experience significant attenuation in tissue, thereby limiting the read range after implantation; e) This effect can be mitigated by employing twice the system power described herein.