I. Core Research Summary

• Ultra-high detection sensitivity, ultra-low detection limit

The sensing detection range spans 0.1–1000 pM, with a limit of detection as low as 9.2 fM and an analytical sensitivity reaching 100 fM—far surpassing that of conventional screen-printed and laser‑etched planar electrodes—enabling the capture of trace amounts of viral antigens.

• Ultra-fast detection in just 43 seconds, with broad-spectrum identification of multiple variants.

After the sample is loaded onto the instrument, complete signal acquisition is achieved in 43 seconds, eliminating the need for prolonged incubation. The assay generates specific impedance responses to all five pseudovirus types—wild-type, Alpha, Beta, Gamma, and Delta—while VSV and the blank control exhibit no significant interfering signals.

• Multi-level 3D micro‑nanostructures significantly enhance capture efficiency.

Aerosol direct‑write gold microcolumns with a high aspect ratio (250 μm in height and 70 μm in diameter), featuring surface‑wrinkled graphene that creates a hierarchical specific surface area, significantly shorten the diffusion path of antigen molecules and accelerate the specific binding between antigens and antibodies.

• High specificity and excellent reproducibility

When high concentrations of interleukins, nucleoproteins, and other interfering proteins are present, signal drift is limited to ±3.5%; the relative standard deviation among four sensors from the same batch is only 2.43%, and storage stability at 4°C for 9 days shows a performance degradation of less than 10%.

• Fully digital maskless additive manufacturing process

Aerosol‑based, site‑specific stacking enables the fabrication of three‑dimensional conductive microstructures without lithography or vacuum deposition; electrode height and array configurations can be rapidly customized on demand, making it well suited for the mass production of miniaturized, portable diagnostic chips.

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

(1) Complete Fabrication Process of the Sensor Chip

• Electron-beam evaporation of Cr/Au seed layers on a glass substrate, followed by laser cutting of the PDMS microfluidic mold;
• 40 wt% gold nanoparticle ink is atomized by ultrasonic atomization, with nitrogen as the carrier gas and a coaxial sheath gas for precise, site-specific deposition.
• Micrometer-scale gold pillar arrays are fabricated via layer-by-layer additive manufacturing, followed by 100°C plate drying and prolonged sintering at 400°C to remove the binder.
• Drop‑coating a reduced graphene oxide dispersion, activating surface carboxyl groups with EDC/NHS, and covalently coupling the SARS‑CoV‑2 spike S1 antibody;
• BSA was used to block non-specific adsorption sites, and the PDMS microfluidic channels were aligned to assemble the detection chip.

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

• Fixed-point self-standing one-step 3D microcolumn fabrication

Traditional planar electrodes are confined to a two-dimensional plane, and etching three-dimensional pillars requires multiple steps of photolithography and deep silicon etching. By contrast, aerosol‑based localized continuous deposition enables the direct fabrication of high aspect‑ratio conductive micropillars without the need for supports or sacrificial layers, allowing multi‑level structures to be formed in a single step.

• Nano–micro–meso hierarchical structure with precise controllability

By freely tuning the printing duration and layer count to control microcolumn height and array spacing, graphene post‑modification is employed to construct a nano‑wrinkled composite interface, enabling the deliberate design of high‑specific‑surface‑area sensing sites and enhancing molecular capture capacity.

• Wide-viscosity nano ink with stable atomization

Compatible with 3 cP low-viscosity, high-solid-content gold nanoparticle ink; ensures long‑term printing without nozzle clogging, uniform droplet deposition, and complete coverage of conductive nanoparticles on the sidewalls and top of micro‑pillars.

• Maskless Digital Rapid Iteration

CAD‑based customization of array layouts and microcolumn dimensions enables rapid prototyping of novel sensor architectures in just a few hours, eliminating the need for mold fabrication or photomask-based patterning and significantly shortening the development cycle of biosensors.

• Compatible with both flexible and rigid substrates, with a compact form factor for easy integration.

Glass, PET, PI, and other substrates are all printable, with an electrode area of only 2×2 mm, making it well-suited to meet the microchip requirements of handheld POCT portable diagnostic devices.

III. Five Key Technological Innovations

Innovation 1: Aerosol-Driven, Site-Specific Stacking of Multi-Level 3D Sensing Microelectrodes

Innovation 2: Integrated Covalent Bio‑Modification Sensing Interface

Innovation: A single-chip, broad-spectrum assay for detecting multiple SARS-CoV-2 variants

Innovation 4: Ultra-Fast Impedance-Based Sensing and Detection System

Innovative 5: A Vacuum-Free, Fully Additive Chip Manufacturing Route

IV. Existing Process Challenges and Corresponding Solutions for Overcoming Them

Challenge 1: Traditional planar electrodes have a low specific surface area, resulting in poor antigen-capture efficiency and a relatively high detection limit.

Challenge 2: Conventional three-dimensional microelectrodes rely on photolithography and etching, resulting in complex processes and extremely high costs.

Challenge 3: Antibodies are prone to desorption through physical adsorption, resulting in poor sensor stability.

Challenge 4: A single chip struggles to distinguish between multiple viral variant samples.

Challenge 5: Conventional printed electrodes require tens of minutes for detection, making it impossible to meet the demands of rapid testing.

V. Future Diversified Industrial Application Scenarios

1. Portable POCT on-site rapid diagnostic chip

2. Pan-Spectrum Screening Platform for Emerging Infectious Diseases

3. High-throughput drug and vaccine evaluation chip for laboratory use

4. Implantable microfluidic biosensor devices

5. Rapid Monitoring Equipment for Environmental Pathogens

Figure 1. Packaging structure of the SARS-CoV-2 volatile organic compound pseudovirus. A) Structure of SARS-CoV-2 and its spike protein; B) Expression of the SARS-CoV-2 spike protein confirmed by Western blotting.

Figure 2. A three-dimensional nanofabricated sensor for the rapid detection of SARS‑CoV‑2 volatile organic compounds. A) Schematic diagram of the three-dimensional nanofabricated sensor. The device comprises a working electrode, a counter electrode, and a reference electrode; the working electrode consists of an array of gold nanoparticle micropillars fabricated using an aerosol‑jet 3D nanoparticle printer. B) SEM image of the nanofabricated micropillar array. C and D) SEM images of the graphene coating on the gold microelectrode array (graphene–gold). The magnified SEM image in D reveals a wrinkled, assembled structure of graphene at the tips of the micropillars. E) Using a unique EDC–NHS coupling chemistry, the surface of the rGO–Au electrode is sequentially functionalized with antibodies targeting the S1 subunit of the SARS‑CoV‑2 spike protein: during the reaction, EDC activates the carboxyl groups on graphene, while NHS serves as a stabilizing agent; an amide bond forms between the spike protein S1 antibody and the graphene–gold surface. After antibodies specific to the SARS‑CoV‑2 spike protein bind to the rGO–Au microelectrode, they inhibit electron transfer caused by oxidation, thereby increasing the electrode impedance.

Figure 3. Sensor characterization. A) Electrochemical impedance spectroscopy (EIS) results show the Nyquist plot of the sensor. The sensor electrode (rGO–Au) was functionalized with SARS‑CoV‑2 spike protein S1 antibodies both before and after measurement. B) Schematic of the equivalent circuit between the counter electrode and the working electrode: Rct denotes the charge-transfer resistance, Rs represents the solution resistance, Cdl corresponds to the double-layer capacitance, and Zw is the Warburg impedance. C) In a typical Nyquist plot, the diameter of the semicircle equals the Rct value. EIS measurements were performed using an AC signal with an amplitude of 1.0 mV and a fixed frequency range of 10,000–1 Hz. D) Cyclic voltammetry (CV) was conducted on the sensor in the presence and absence of spike protein S1 antibodies; upon antibody coating on the electrode surface, the sensor current decreases due to the electron‑blocking effect of the antibodies. All EIS and CV measurements were carried out in a solution containing PBS buffer (50 mM) and the redox mediator ferricyanide/ferrocyanide (5 mM).

Figure 4. Calibration study of the SARS‑CoV‑2 sensor. A) Detection results for SARS‑CoV‑2 spike protein S1 antigen at different concentrations. Using electrochemical impedance spectroscopy (EIS), the relationship between S1 antigen concentration and the Nyquist plot was determined over the range of 0.1–1000 pM; B) Plots showing how the corresponding Rct values vary with different S1 antigen concentrations. In the Nyquist plots, the deviation between the fitted semicircle (see Figure 3C) and the original semicircle is less than ±3.0%; a p‑value < 0.0001 indicates a statistically significant difference (B); C) The variation of the sensor’s imaginary impedance (Zim) with measurement time (seconds) at different antigen concentrations; D) The dependence of the impedance modulus (|Z|) on frequency across various concentrations. Each antigen concentration was measured in triplicate. All EIS experiments were conducted using an AC signal with an amplitude of 1 mV and a frequency sweep from 10,000 to 1 Hz. For antigen titrations, PBS (50 mM) containing ferrocyanide/ferrocene (5 mM) was used as the redox mediator.

Table 1. Comparison of the performance of our 3D-printed biosensor in detecting the SARS-CoV-2 spike protein with results reported in the literature.

Figure 5. Detection of SARS-CoV‑2 volatile organic compounds. Based on the titration results shown in Figure 4, this sensor was used under identical conditions to quantify SARS‑CoV‑2 VOCs. A) Different methods for preparing SARS‑CoV‑2 pseudoviruses are illustrated in Figure 1, and these were tested using EIS measurements with the sensor; in this study, VSV and Mock served as controls, while WT, Alpha, Beta, Delta, and Gamma represented the five VOCs examined. B) Rct values corresponding to the control group and each VOC were calculated, and plots were generated to assess p-values (<0.0001); each virus and the control group were measured at least three times. C) For each EIS measurement, curves depict the relationship between the detection phase and impedance modulus as a function of frequency. D) Curves show how the Zim values of sensors detecting different VOCs vary with measurement time (in seconds). EIS measurements for antigen detection were performed by applying an AC signal with an amplitude of 1 mV and a frequency range of 10,000–1 Hz; experiments employed PBS (50 mM) containing iron/ferrocyanide (5 mm) to prepare recombinant spike protein S1 antigens at various concentrations.

Figure 6. Selectivity and repeatability of the sensor. In this study, the sensor’s selectivity was evaluated under conditions containing or lacking SARS‑CoV‑2 spike protein antigen (0.1 pM), in the presence of various interfering antigens, including interferons, IL‑6, and non‑cytokine proteins. The concentration of each interfering antigen was set at 10 pM, while the sensor’s baseline response was defined as the level observed at an SARS‑CoV‑2 S1 antigen concentration of 0.1 pM. Subsequently, each interfering antigen was added to the solution at a 1:1 molar ratio. A) Nyquist plots illustrating the sensor’s behavior under different interfering antigen conditions; B) Corresponding Rct values calculated by fitting the impedance data from panel A; C) Nyquist plots obtained from four identical sensors tested in the same buffer containing SARS‑CoV‑2 antigen (0.1 pM); D) The corresponding Rct values. Compared with the baseline response, the p-values for all sensors were less than 0.0001. The experiments were conducted using a PBS (50 mM) solution containing ferrocyanide (5 mM) as the redox mediator.