I. Core Research Summary

• Multilayer Integrated All-Printed Aerosol Architecture

A single device stack alternately deposits a polyimide (PI) insulating layer, silver conductive traces, and an FCG graphene protective layer/sensing electrode. The structure is then transferred onto a silicone elastomer substrate via a PMMA sacrificial layer, resulting in a total device thickness of less than 2 mm and a weight of under 5 g. It conforms seamlessly to the curvature of the arm and can withstand 180° bending and 60% biaxial stretching.

• Dual-core material advantages: biocompatible + antioxidant and solderable

High aspect ratio FC graphene exhibits no cytotoxicity, with keratinocyte viability on it comparable to that on standard culture dishes after 7 days. Its silver‑line surface coating provides complete oxidation resistance, preventing solder from attacking the silver pads during reflow soldering and thereby addressing the challenges of chip mounting on flexible circuits.

• Gel-free, high-fidelity electromyography acquisition

The FCG mesh electrode exhibits excellent skin–electrode contact, achieving a signal-to-noise ratio comparable to that of medical‑grade gel electrodes even without gel. It demonstrates minimal signal drift during stretching and movement, thereby avoiding issues associated with gel drying and skin sensitization.

• Long-distance stable wireless flexible circuit

All‑print flexible Bluetooth antenna delivers signal strength comparable to that of a rigid PCB within 15 meters, with resistance variation of less than 1% after 100 bends, ensuring long‑term circuit stability.

• Multi-channel intelligent gesture recognition

Single-electrode 4‑class gesture recognition achieves 99% accuracy; a three‑electrode configuration distinguishes between open/close hand gestures and seven independent five‑finger motions, with an accuracy of 98.6%, enabling wireless remote control of drones, robotic arms, and presentation software.

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

(1) Complete multi-layer flexible electronic printing process

• Spin-coat a PMMA sacrificial layer onto the glass substrate, followed by plasma surface activation.
• Pneumatic atomization PI insulating ink, printed onto the underlying support film, cured at 250°C;
• Ultrasonic atomization of silver nanoparticle ink, printing of underlying conductive traces, followed by high‑intensity pulsed light sintering.
• The PI dielectric layer is again printed via aerosol jet printing, with vias left open to enable interlayer connectivity.
• Ultrasonic atomization FC graphene ink is used to print the top-layer antioxidant protective coating and the sensing electrodes.
• The outer PI encapsulation is removed, the sacrificial layer is dissolved in acetone, and the complete circuit is transferred onto flexible Ecoflex silicone.
• Low-temperature solder paste, surface-mount Bluetooth modules, operational amplifiers, miniature chips, and miniature lithium batteries are assembled.

(II) Five Core Process Advantages (Compared with Lithography, Sputtering, and Screen Printing)

• Dual-atomization compatibility with multi-viscosity inks

The aerosol system is equipped with both pneumatic and ultrasonic atomization units, enabling simultaneous fabrication of high-viscosity PI insulating adhesive, medium-viscosity silver conductive paste, and low-concentration graphene aqueous dispersion. A single production line can complete the entire stack—conductive, insulating, and sensing layers—eliminating the need to switch between multiple pieces of equipment.

• Micron-level precision through-hole interconnects, with multi-layer circuits integrally formed.

Digital processing enables precise placement of vias; the PI insulating layer is pre‑patterned with 50 μm conductive vias, and the upper and lower silver traces are automatically interconnected, eliminating the need for photolithographic window opening and wet etching, thus achieving multi‑layer wiring in a single step.

• Gentle deposition at room temperature, compatible with biofunctional materials.

Without any high-temperature, high-pressure forming steps, the graphene-based biomaterials do not undergo thermal degradation, fully preserving their skin-friendly properties and making them well-suited for the mass production of medical wearable devices.

• On-demand localized additive manufacturing significantly reduces material costs.

Nano‑ink is deposited only in the circuit areas, reducing material loss from full‑area sputtering of gold and silver by 70% and enabling rapid prototyping and iterative development of small‑batch gesture‑sensing samples.

• Flexible, compatible transfer process, compatible with various elastic substrates.

After printing, the sacrificial layer is dissolved away, enabling the intact nanofilm to be peeled off without damage onto silicone or PI flexible substrates, allowing direct fabrication of ultra-thin, skin‑like electronic devices.

III. Five Core Technological Innovations of This Study

Innovation 1: A solderable, oxidation-resistant functional graphene (FCG) ink system

Innovation 2: In-situ Multilayer Three-Dimensional Wiring Process for Aerosols

Innovative 3D Gel-Free Skin-Conformal Graphene Sensing Electrode

Innovative 4‑layer sacrificial transfer aerosol nanofilm technology

Innovative 5‑in‑1 aerosol printing hardware and machine learning–integrated interaction solution

IV. Existing Process Challenges and Corresponding Solutions for Overcoming Them

Challenge 1: Silver nano‑line interconnects are highly susceptible to oxidation and corrosion during high‑temperature reflow soldering, leading to pad delamination.

Challenge 2: Direct contact between the silver electrode and the skin leads to the release of metal ions, which can induce cytotoxicity.

Challenge 3: Traditional printed electrodes are extremely dense and smooth, making them prone to generating motion-induced noise during bodily movements.

Challenge 4: Multi-layer printing is prone to interlayer misalignment and via open circuits.

Challenge 5: In flexible circuits, repeated bending can easily cause the conductive traces to fracture.

V. Future Diversified Industrial Application Scenarios

1. Rehabilitation Medical Myoelectric Prosthesis Control System

2. VR/AR Gesture-Based Human-Computer Interaction Device

3. Real-time athletic performance monitoring equipment

4. Portable medical screening device

5. Industrial Remote Wireless Control

Figure 1. Design, structure, and nano‑fabrication process of the p‑NHE. a, b: Schematic illustrations of aerosol nanoparticle printing using PI, Ag, and graphene inks to construct multilayer architectures. After atomization in an N₂ gas stream, droplets are directed toward the print head to form the desired pattern on the substrate (a); the aerosol flow is sheathed and guided to the substrate; in this study, three inks with distinct viscosities were printed sequentially (b). Insets show partial chemical structures of oxidized graphene bearing carboxyl and hydroxyl groups. c: Photographs depicting the continuous printing process of stretchable nanomembrane electrodes using PI and graphene inks: Step 1, PI ink is printed onto a PMMA/glass substrate as a support layer; Step 2, graphene ink is printed on the PI surface as the conductive electrode layer. d: Static images of the multilayer bio‑potential recording circuit fabrication process: Step 1, Ag ink is deposited on the PI/PMMA/glass substrate as the circuit’s ground layer; Step 2, PI ink is printed over all areas except the VIA hole locations to serve as the dielectric layer; Step 3, Ag and FCG inks are printed to form the metal interconnect layer connected to the ground plane. In Step 4, additional PI ink is applied to fully encapsulate the entire circuit, excluding the contact pads for integrating functional chip components. e, f: Cross‑sectional profiles of the printed stretchable electrode (e) and the printed multilayer circuit (f). g, h: Close‑up photographs showing the electrodes (g) and the circuit (h) from e and f mounted on an elastic membrane.

Table 1: Comparison of Multilayer Sensors and Circuit Systems

Figure 2. Preparation and characterization of FCG, along with its application as a sensor. a Atomic force microscopy image (top view) of FCG initially dispersed in an aqueous ink on a mica substrate. b Three height distribution profiles of FCG flakes, corresponding to the three dashed lines in panel a. c and d show low‑resolution (c) and high‑resolution (d) transmission electron microscopy images of FCG, respectively. e Schematic illustrating the stacked structure of printed FCG before and after droplet drying. f Color‑coded cross‑sectional scanning electron microscopy image revealing a multilayer sensor architecture comprising graphene, poly(methylenediphenylamine) (PI), and an elastomeric substrate. g Enlarged view of panel f, depicting the FCG stack with a total thickness of 800 nm. h Electromyographic signals recorded from the forearm, comparing the signal amplitudes of four electrode types: printed FCG (red), gold–green, silver–blue, and commercial metal/gel (black). i Root-mean-square values of the electromyographic signals shown in panel h. j Average signal‑to‑noise ratios for printed FCG (red), gold–green electrodes (green), silver electrodes (blue), and commercial gel electrodes (black); error bars indicate standard deviation (n = 3). k Optical images demonstrating the layered adhesion and stretchability of FCG electrodes on skin: longitudinal stretching (top) and compression (bottom) of the forearm. l Optical microscopy images of cultured keratinocytes exposed to pristine DMEM (left; control sample) and to DMEM supplemented with graphene after 7 days of co‑culture (right). m and n: comparison of absorbance (m) and fluorescence intensity (n) of keratinocytes cultured on five different membrane surfaces, including a control group, an elastomeric membrane, a graphene‑coated membrane, a gold‑coated membrane, and a silver‑coated membrane. Overall, these data clearly demonstrate that graphene electrodes foster a cell‑friendly environment when integrated into elastomeric membranes.

Figure 3. Printing, integration, and characterization of multilayer electronic circuits. a Schematic illustrating the structure of a multilayer‑printed flexible wireless electronic device; b FIB‑assisted SEM images showing the cross‑section of the multilayer circuit printed along the direction indicated in (a), with a platinum layer deposited on the circuit surface to protect its structure during cross‑section imaging; c and d Comparison of the solderability of two materials: the left‑hand schematic depicts circuits containing only a silver layer (c) and a silver layer coated with FCG (thickness 0.1 µm) (d), the latter demonstrating that FCG effectively prevents flux‑induced corrosion of the silver layer; the central and right‑hand photographs clearly highlight the difference: without FCG, the silver‑based contact pads are removed during the reflow soldering step; e and f Characterization of sintered silver layers—both with and without printed FCG—using XRD (e) and XPS (f): XRD data (e) confirm that the printed FCG provides oxidation resistance, while XPS data (f) reveal that the FCG/silver film exhibits diffraction peaks only at 38.0°, 44.1°, and 54.8°, indicating a pure silver structure with no oxidation; g Resistance‑change curves for a highly flexible printed circuit after 180° bending (up to 100 cycles), showing negligible resistance variation. The inset photograph illustrates the test circuit being wrapped around the edge of a bent glass substrate to achieve 180° curvature; h This figure simultaneously compares the RSSI response characteristics of commercial rigid circuits and printed circuits, demonstrating that the printed circuit delivers comparable performance when operating over Bluetooth‑based communication distances of up to 20 meters.

Figure 4. Illustrates a wearable p‑NHE system for wireless, multi‑class human–machine interaction. (a) The schematic diagram (top) shows the target muscles on the forearm used to recognize multiple gestures; the bottom image depicts three p‑NHE devices placed on the palmaris longus, brachioradialis, and flexor carpi ulnaris, among other target muscles. Enlarged views reveal the system’s circuitry and electrode configuration. (b) EMG mapping data present multi‑channel root‑mean‑square signals covering the entire forearm; six heatmaps are generated—corresponding to hand closure, thumb, index finger, middle finger, ring finger, and little finger movements (from left to right)—to identify optimal channel locations. (c) Typical EMG signals for four gestures: open hand, closed hand (left), index finger flexion, and wrist flexion (right). (d) Demonstrates an electromyography‑driven human–machine interaction system implemented with a wireless, wearable p‑NHE device, enabling precise control of a flying drone (left), a remote‑controlled vehicle (center), and demonstration software (right). In this experiment, a single‑channel device was mounted on the palmaris longus muscle (as shown in Figure a). (e) Displays a three‑dimensional distribution map of three‑channel EMG root‑mean‑square signals, clearly distinguishing seven distinct gestures and assigning them to seven categories. (f) Summarizes real‑time confusion matrix results from ten trials, showing that three‑channel EMG recording achieves 98.6% accuracy in recognizing the seven gesture classes. (g) Illustrates the correspondence between three‑channel EMG recordings and robotic hand‑motion control, presenting six example movements performed by the subject along with the resulting trajectories of the robot’s hand.