Non-targeted single-molecule conductivity spectrum measuring device, device and method

Abstract

The invention provides a non-targeted single-molecule conductivity spectrum measuring device, device and method, which are based on a tunneling device of integrated dielectrophoresis tweezers, and in a complex matrix represented by urine, molecules are subjected to extremely-low-concentration dielectrophoresis capture and enrichment and tunneling electrical signal detection, namely consubstantial capture and detection. And the detection target does not need to be predefined before detection. Through the strong capturing and enriching capability of the dielectrophoresis tweezers, the collection of various molecular information in the urine matrix is realized. Tunneling electrical signal spectrums corresponding to different urine samples can be obtained by using the device, then multi-dimensional feature extraction is performed on the signal spectrums, and accurate classification of different samples is realized through a machine learning model.

Description

Technical Field

[0001] This invention relates to the field of single-molecule detection and micro / nano device technology, specifically to a non-targeted single-molecule conductivity spectroscopy measurement device, apparatus, and method. Background Technology

[0002] Early diagnosis of disease is crucial for prognosis. Currently, detecting molecular markers in complex matrices such as urine has proven to be an effective approach for early disease diagnosis, a key requirement in modern clinical diagnostics, life science research, and environmental monitoring. However, these matrices are extremely complex, containing high concentrations of electrolytes, proteins, metabolites, and cellular debris, posing significant challenges to the sensitivity, specificity, and interference resistance of detection technologies.

[0003] Currently, among routine detection and analysis methods, physicochemical and morphological analysis serves as a fundamental clinical diagnostic tool. Automated biochemical analyzers are used to detect sample pH, osmotic pressure, and electrolyte concentration. Simultaneously, flow cytometry or microscopy is used to count and observe formed elements such as cells, casts, and crystals, providing macroscopic indicators for preliminary diagnosis. Immunological detection techniques, based on the principle of specific antigen-antibody recognition, encompass classic techniques such as enzyme-linked immunosorbent assay (ELISA), chemiluminescent immunoassay (CLIA), colloidal gold lateral chromatography, and electrochemiluminescence (ECL), and are widely used for the quantitative detection of specific proteins, hormones, or drug molecules in complex matrices. Chromatography-mass spectrometry (GC-MS) is an advanced method for analyzing components in complex matrices. Gas chromatography (GC) or liquid chromatography (LC) separates components based on their polarity, molecular weight, and other physicochemical properties, followed by precise identification by mass-to-charge ratio using mass spectrometry (MS). This is a core method in metabolomics and proteomics research. Nucleic acid amplification and analysis techniques target trace amounts of nucleic acids (such as cfDNA and miRNA) in samples. (e.g.) Using polymerase chain reaction (PCR), isothermal amplification technology or next-generation sequencing (NGS) for sequence analysis and quantification, providing support for the diagnosis of gene-related diseases; spectral analysis technology utilizes the selective absorption, scattering or emission characteristics of molecules to electromagnetic waves, and uses techniques such as ultraviolet-visible absorption spectroscopy (UV-Vis), infrared spectroscopy (IR) and fluorescence spectroscopy to evaluate the overall chemical composition of the matrix or the distribution of specific functional groups.

[0004] Single-molecule conductivity spectroscopy is an ultra-sensitive characterization technique that studies electronic transport properties at the single-molecule level. It has significant advantages such as being label-free, having a fast response speed, a high signal-to-noise ratio, and being able to directly reveal the essential energy level structure of molecules through electronic transport fingerprints. Its core is to construct nanoscale metal electrode gaps and record the electrical signals when molecules pass through or bridge the gaps. Representative technical paths include scanning tunneling microscopy split junction technology (STM-BJ), mechanically controlled split junction technology (MCBJ), and single-molecule electric field-effect transistor (Single-Molecule FET). Among them, Scanning Tunneling Microscopy-Binding Technology (STM-BJ) uses the probe of a scanning tunneling microscope to repeatedly perform "contact-stretch" cycles on the substrate, capturing the conductivity signal of the analyte molecule at the moment the electrode is about to break, and obtaining the conductivity fingerprint spectrum of the molecule through statistical analysis; Mechanically Controlled Binding Technology (MCBJ) controls the micro- and nano-mechanical displacement of metal nanowires on a flexible substrate, causing them to break through by bending and stretching to create atomic-level gaps. This method has extremely high mechanical stability, can maintain the single-molecule detection environment for a long time, and can carry out precise conductivity spectrum measurements; Single-Molecule FET uses a single molecule as a semiconductor channel connected between the source and drain electrodes, modulating the energy level of the molecular orbital by the gate voltage, thereby detecting the electrical transport characteristics of the molecule and its charge response to the external environment.

[0005] While the aforementioned technologies have certain applications in their respective fields, they still have significant shortcomings and deficiencies in achieving direct, rapid, single-molecule conductivity spectroscopy analysis of extremely low abundance biomarkers in complex matrices. In conventional detection and analysis methods, spectroscopic detection techniques typically have high detection limits, while immunological detection (such as ELISA and CLIA), although possessing high sensitivity, heavily relies on the specific binding of exogenous antigens and antibodies. This not only results in high reagent costs and susceptibility to interference from non-specific cross-reactions in the matrix, but also only provides macroscopic concentration information, failing to obtain the intrinsic electronic transport characteristics (conductivity fingerprint) of the target analyte at the single-molecule level. Mass spectrometry coupled with chromatography, while accurate, requires extremely high sample purity, necessitating complex offline pretreatment processes such as desalting, protein precipitation, and solid-phase extraction. This is not only time-consuming and labor-intensive but also highly susceptible to loss or denaturation of low-abundance target components during processing. Physicochemical analysis and conventional biochemical detection can only provide concentration or macroscopic morphology information, failing to obtain the electronic transport characteristics (conductivity fingerprint) or energy level structure of the target analyte at the single-molecule level, making it difficult to accurately distinguish isomers or subtle structural differences.

[0006] Existing single-molecule conductivity spectroscopy detection technologies also suffer from numerous problems. Technologies such as STM-BJ and MCBJ primarily rely on the random diffusion of molecules into the nanoelectrode gaps due to Brownian motion in solution. In complex matrices such as urine, the concentration of target molecules is extremely low, and there is significant background molecule interference. Relying solely on passive diffusion results in a very low probability of target molecules entering the detection region, leading to extremely long waiting times for single effective detections and failing to meet the needs of rapid clinical testing. High concentrations of electrolytes and non-target proteins in complex matrices can severely interfere with detection. For tunneling detection (STM-BJ / MCBJ), impurity molecules easily undergo non-specific adsorption on the electrode surface, leading to electrode contamination and... False conductivity signals are generated, and high ion concentrations significantly increase background leakage current, drowning out weak quantum tunneling signals. For field-effect transistors (FETs), solution ions in physiological saline matrix form a tight electric double layer on the gate surface, producing a severe Debye shielding effect. This compresses the sensing distance of the device to the change in the charge of the target molecule to below nanometers, leading to a sharp drop in detection sensitivity or even failure. In addition, existing devices such as mechanically controlled crack junctions (MCBJ) usually require a large shockproof system and a precise piezoelectric control platform. The devices are large in size and costly, and it is difficult to integrate them on-chip with microfluidic sample enrichment modules, making it impossible to achieve integrated "sampling-enrichment-detection" operation.

[0007] In existing technologies, not only is there a lack of active enrichment mechanisms, but the electrode surface also lacks anti-adsorption design. Impurities in complex matrices are easily adsorbed non-specifically on the electrode surface, leading to electrode contamination and the generation of false signals.

[0008] In summary, existing technologies struggle to overcome the high background noise interference of complex matrices and achieve effective capture and single-molecule conductivity spectroscopy characterization of extremely low-abundance biomolecules without requiring complex pretreatment. There is an urgent need for an efficient, sensitive, and integrated single-molecule conductivity spectroscopy testing device and analysis method suitable for complex matrix environments. Summary of the Invention

[0009] To address the aforementioned shortcomings of existing technologies, this invention proposes a non-targeted single-molecule conductivity spectroscopy measurement device, apparatus, and method. This method enables the capture and enrichment of molecules at extremely low concentrations using dielectrophoresis in complex matrices, such as urine, and the detection of tunneling electrical signals—essentially, in-situ capture and detection. No pre-definition of the detection target is required before detection. The strong capture and enrichment capabilities of dielectrophoresis tweezers allow for the collection of information from multiple molecules within the urine matrix. Using this device, tunneling electrical signal spectra corresponding to different urine samples can be obtained. Subsequently, multi-dimensional feature extraction is performed on the signal spectra, and a machine learning model is used to accurately classify different samples.

[0010] On one hand, the present invention provides a non-targeted single-molecule conductivity spectrum measurement device, characterized in that it includes a single-molecule enrichment unit and a tunneling detection unit; the single-molecule enrichment unit is used to capture, enrich or locate single molecules in the sample to be tested to or near the nano-gap; the tunneling detection unit is used to perform non-targeted detection of single molecules entering the nano-gap; the single-molecule enrichment unit has nanoscale control precision, including any one or more of dielectrophoresis, electrophoresis, and electrochemical tweezers.

[0011] This invention integrates the active enrichment and spatial manipulation of target molecules, as well as the in-situ acquisition of single-molecule conductance signals based on the quantum tunneling effect, through the same probe structure.

[0012] Furthermore, the single-molecule enrichment unit is a dielectrophoresis manipulation unit, including 1 to 4 auxiliary electrodes; the tunneling detection unit includes a pair of nano-gap electrode pairs; the gap width of the nano-gap electrode pairs is sub-5 nanometers, and the surface is modified with a cysteine ​​or mercaptoethanol anti-adsorption molecular layer; the auxiliary electrodes and the nano-gap electrode pairs are arranged orthogonally, nested or adjacent in space.

[0013] Unlike the passive diffusion mechanism relying on thermal Brownian motion in traditional fixed-gap electrode systems (i.e., molecules are limited by random thermal motion and enter the detection region with low probability through random collisions, resulting in severe diffusion restrictions and uncontrollability), this invention utilizes dielectrophoretic manipulation units to construct a long-range directional mass transport mechanism from the bulk solution to the nano-gap. This method generates a high-intensity electric field gradient at the tunneling gap by applying an AC signal with specific parameters to the auxiliary electrode, thereby overcoming the fluid resistance and thermal disturbance in the solution and directionally migrating the target molecule from the far-field region and highly localizing it to the electric field maxima region of the nano-gap. This not only effectively solves the mass transfer efficiency bottleneck in the nano-confined space, but also achieves deterministic control of the rate and orientation of single molecules entering the detection region by precisely modulating the electric field frequency, amplitude, and application timing.

[0014] Dielectrophoresis relies on a non-uniform strong electric field to manipulate particles. However, the tunneling electrodes have extremely small spacing and extremely high surface electric field strength and gradient, which can exert a significant dielectric attraction on charged or polarizable biomolecules and impurities in the solution. This causes them to accumulate on the electrode surface and generate strong non-specific adsorption, which seriously interferes with the tunneling signal and detection specificity. Therefore, the electrode surface must be modified to be hydrophilic and antifouling to suppress adsorption.

[0015] In some embodiments, this invention modifies only the tunneling electrode to prevent adsorption, while the dielectrophoresis-assisted electrode retains its original conductive surface. This solves the problem of non-specific adsorption in the tunneling detection zone and avoids significant interference from the modified layer on the generation and propagation of the high-voltage AC field in dielectrophoresis, achieving a dual optimization of "enrichment efficiency" and "detection anti-fouling properties." Cysteine ​​and mercaptoethanol can both form a dense, self-assembled monolayer on the electrode surface: the thiol groups in their molecular structure form strong Au-S covalent bonds (gold-sulfur bonds) with the carbon-based gold composite electrode surface, thereby sealing the active sites on the electrode surface and physically blocking non-specific adsorption. Simultaneously, the exposed terminal amino and carboxyl groups of cysteine ​​carry a charge under specific pH conditions, forming a hydrophilic and charged interface. This electrostatic repulsion prevents molecules with the same charge in the solution from approaching each other. This dual effect of physical sealing and chemical repulsion jointly constructs a clean and anti-fouling electrode interface.

[0016] The auxiliary electrode pair and the nano gap electrode pair are not independent but are integrated on the same probe or chip substrate and form a specific spatial arrangement. This arrangement enables the dielectric electric field generated by the auxiliary electrode to directly drive molecules in the solution and accurately locate them to the detection hotspot region of the nano gap electrode pair, thereby realizing the active capture, localization and signal acquisition of single molecules.

[0017] Preferably, using a pair of auxiliary electrodes yields the best results.

[0018] The adjacency relationship between the auxiliary electrode and the nanogap electrode pair is such that the tip regions of the two pairs of electrodes are close to each other in space, but their physical structures are completely separated. For example, if they are fabricated side by side on the same substrate, the spacing may be on the order of hundreds of nanometers to micrometers. The orthogonal relationship is such that the two pairs of electrodes are perpendicularly intersecting in three-dimensional space, and their center points coincide in space, but the electrode entities themselves may be on different planes and isolated by a dielectric layer. The nested relationship is a deeper form of the orthogonal relationship, in which the tip of the auxiliary electrode approaches the outer opening of the tunnel junction of the main electrode in three-dimensional space, forming a local envelope or microcavity structure, maximizing the focusing effect of the electric field gradient on the gap interior.

[0019] Preferably, the auxiliary electrode and the nano-gap electrode pair are orthogonal in space, which is a preferred balance between performance and feasibility.

[0020] The device further includes an insulating substrate, through which the dielectric electrophoresis manipulation unit and the tunneling detection unit are insulated from each other; the insulating substrate is made of any one or more of quartz, glass tube, silicon wafer, glass sheet, polymer, and flexible material.

[0021] In this device, a high-frequency AC voltage of 1 V to 15 V is applied to the auxiliary electrode to generate dielectric force, while the nano-gap electrode measures a weak DC tunneling current in the picoampere (pA) range. The insulating substrate ensures that there is no leakage or short circuit between these two sets of high-voltage electrodes and the weak signal detection electrode, preventing strong AC signals from directly entering the high-sensitivity current amplifier, which could lead to signal overload or device damage. By etching, depositing, or pulling on the insulating material, the electrode positions can be precisely formed to be mutually insulated but spatially fixed.

[0022] In another aspect, the present invention provides a non-targeted single-molecule conductivity spectrum measurement device, including the measurement device, microfluidic channel sample cell and support as described above, and an electrical control module.

[0023] The microfluidic channel sample cell and support structure provide the device with a controllable, stable, and micro-nano scale reaction and detection environment that matches the core components; the electrical control module provides the device with precise "power" and "sensing", namely generating the manipulation field, collecting weak signals, and completing preliminary processing.

[0024] Furthermore, the microfluidic channel sample cell includes a glass tube with an inner diameter of 1.5 mm to 2.5 mm and its support frame; the support frame tilts the glass tube at an angle of 10 to 30 degrees; it also includes an electrical control module, which includes a working electrode, a reference electrode, a function signal generator, a patch-clamp amplifier, and a digital-to-analog converter.

[0025] The inner diameter of the glass tube is 1.5 mm to 2.5 mm, an optimized result that balances hydrodynamic performance, ease of use, and manufacturing cost. If the inner diameter is too small, the fluid resistance within the microfluidic channel increases dramatically, leading to difficulties in sample introduction and making the channel highly susceptible to blockage by complex matrices (such as urine containing crystals or cell debris). Simultaneously, under the same injection pressure, the extremely high shear stress within the confined space can easily damage fragile biological samples or alter their natural conformation. If the inner diameter is too large, the flow velocity within the tube decreases significantly, easily creating fluid eddies or "dead volumes" in dead zones, preventing efficient and uniform sample transport to the detection tip. Furthermore, the increased channel space reduces the mass transfer efficiency of the sample on a macroscopic scale, significantly increasing the time required for target molecules to reach the enrichment area at the probe tip via convection and diffusion, severely weakening the target capture rate per unit time. These hydrodynamic mass transfer delays and losses, combined, ultimately lead to a significant decrease in the overall detection efficiency and signal response time of the system.

[0026] In another aspect, the present invention provides the use of the above-mentioned measuring device or measuring apparatus in non-targeted detection of complex matrices.

[0027] On another aspect, the present invention provides a method for measuring non-targeted single-molecule conductivity spectra. The method uses the above-mentioned measuring device or measuring apparatus for detection, and contacts the sample to be tested with the measuring device in a microfluidic channel sample cell to obtain conductivity spectroscopic characterization, thereby achieving non-targeted characterization of the in-situ, real-time electronics of single molecules in the sample to be tested.

[0028] After the sample is injected into the microfluidic channel, it is fixed to the detection area by the support. The electrical control module drives the auxiliary electrode to generate a dielectric field, which actively enriches the target molecules near the nano gap. The module switches to the detection mode, applies a bias voltage to the nano gap, and acquires the tunneling current signal with high fidelity through an amplifier and a digital-to-analog converter. Finally, subsequent conductivity spectrum analysis, feature extraction and machine learning classification are performed.

[0029] Furthermore, the method includes the following steps: (1) The sample to be tested and the measuring device are brought into contact in the microfluidic channel sample cell, the tunneling detection unit is kept energized, the dielectrophoresis manipulation unit is turned on to apply an alternating electric field, and the molecules in the sample are driven and enriched into the detection gap. (2) Turn off the dielectrophoresis manipulation unit, and obtain the non-targeted single-molecule conductivity spectrum of the sample to be tested through the tunneling detection unit. Repeat steps (1) and (2) more than 4 times to maintain the molecular capture probability in the detection gap and accumulate single-molecule conductivity spectrum data. (3) After analysis and feature extraction, a model is constructed to classify the samples to be tested.

[0030] Further, the peak-to-peak value of the sinusoidal AC signal of the dielectrophoresis manipulation unit is 1~20 Vpp, and the frequency is 100kHz~5MHz; the time for applying the AC electric field in step (1) is 5~20 s; the bias voltage applied to the tunneling detection unit in step (2) is 50~300 mV, and the collection time for obtaining the non-targeted single-molecule conductivity spectrum of the sample to be tested is 4~10 min.

[0031] Furthermore, the features include feature parameters extracted from multiple dimensions such as time-domain statistical features, Fourier spectrum features, power spectrum features, wavelet features, and nonlinear features; the test sample is any one or more of urine, serum, saliva, milk, or wine; the classification of the test sample includes any one or more of the following: differentiation between disease and health, differentiation between different types of disease, pathogen identification, and cancer subtyping.

[0032] This invention can be applied to clinical diagnosis (such as serum and urine) and food safety and quality monitoring (such as milk and wine). The method has the ability to work directly in physiologically relevant complex fluid environments such as low ionic strength solutions, serum or urine. It can continuously and label-free collect electronic signals of metastable intermediates in the process of biomolecular interaction, thereby resolving the randomness and heterogeneity masked by macroscopic population statistical averaging effects at the single-molecule scale. It has the technical advantages of high-throughput screening and low-cost, rapid and high-precision detection.

[0033] The present invention has the following beneficial effects: (1) Active capture of single molecules: The active enrichment effect of dielectrophoresis overcomes the limitation of traditional single molecule technology that relies on passive diffusion. Combined with the label-free characteristics of quantum tunneling detection, an integrated "enrichment-detection-analysis" technology path suitable for complex matrices such as urine has been successfully constructed.

[0034] (2) High sensitivity: By combining active capture by dielectrophoresis with direct measurement by tunneling, the detection limit is increased from nanomolar level to femtomolar level, solving the problem of detecting molecules with extremely low abundance.

[0035] (3) Simple pretreatment: Relying on physical principles (locality of tunneling and screening properties of dielectrophoresis), it can directly analyze complex matrix samples such as urine, saving time-consuming purification steps and achieving rapid in-situ detection.

[0036] (4) Provides novel molecular conductivity spectra: It obtains not only concentration, but also single-molecule conductivity spectra containing information such as molecular structure and conformation. It can distinguish minute structural differences without pre-specifying targets and is suitable for the discovery of unknown substances.

[0037] (5) Compared with traditional methods that only detect one or more biomarkers (ELISA, etc.), the information obtained by the scheme of detecting a broad spectrum of biomarkers in the sample is richer.

[0038] (6) Flexible application scenarios: The method can be extended to other complex matrices similar to human urine, such as wine, milk or other body fluids.

[0039] (7) Anti-adsorption contamination: The cysteine ​​molecular layer modified on the surface of the tunneling detection electrode can significantly reduce the non-specific adsorption of impurities in complex matrices, reduce electrode contamination and false signals, and improve the detection stability and reusability of the device. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the non-targeted single-molecule conductivity spectroscopy measurement device in Example 1; Figure 2 This is a schematic diagram (cross-section) of the non-targeted single-molecule conductivity spectroscopy device in Example 1. Figure 3 This is a schematic diagram of the non-targeted single-molecule conductivity spectroscopy device in Example 1; Figure 4 SEM characterization of the tip of the four-hole quartz capillary after it was drawn in Example 1 (left) and optical microscope characterization of the tip of the device (right). Figure 5 This is a flowchart illustrating the detection process using the non-targeted single-molecule conductivity spectroscopy measurement device in Example 2. Figure 6 The tunneling electrical signal spectrum of the urine sample in Example 2 (Sample 1); Figure 7 The tunneling electrical signal spectrum of the urine sample in Example 2 (Sample 2); Figure 8 The tunneling electrical signal spectrum of the urine sample in Example 2 (Sample 3); Figure 9 The tunneling electrical signal spectrum of the urine sample in Example 2 (Sample 4); Figure 10 The tunneling electrical signal spectrum of the urine sample in Example 2 (sample 5); Figure 11 This is a spatial arrangement diagram of the electrophoresis and nano-gap electrode pairs in Example 4 (in order: orthogonal, adjacent, nested). Figure 12 This is a schematic diagram of different numbers of electrodes in Example 5 (1, 2, and 4 electrodes in sequence). Figure 13 The electrical signal spectrum of the device in Example 6 for detecting extremely low concentrations of gold nanoparticles; Figure 14 This is the electrical signal spectrum of the device in Example 6 for detecting extremely low concentrations of protein. Detailed Implementation

[0041] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be noted that the embodiments described below are intended to facilitate understanding of the present invention and are not intended to limit it in any way. The reagents used in this embodiment are all known products, and unless otherwise specified, they are all commercially available products.

[0042] Example 1: A non-targeted single-molecule conductivity spectroscopy measurement device The non-targeted single-molecule conductivity spectroscopy measurement device provided in this embodiment is as follows: Figure 1As shown, Identifier 1 is a dielectrophoresis-tunneling single-molecule conductivity spectroscopy detection device based on a four-hole quartz capillary; Identifier 2 is an electrical signal acquisition module, including a patch-clamp amplifier and a digital-to-analog converter module, where Identifier 4 is the working end of the device and Identifier 5 is the grounding end of the device; Identifier 3 is an electrical control module, including a function signal generator; Identifier 6 is the microfluidic channel sample cell of the system; and Identifier 7 is the Ag / AgCl electrode used as a reference electrode.

[0043] The fabrication process of the dielectrophoresis-tunneling single-molecule conductivity spectroscopy detection device 1 based on a four-hole quartz capillary is as follows: (1) Pretreatment of quartz capillary tubes Four-barrel quartz capillary tubes with an outer diameter of 1.5 mm and an inner diameter of 1.10 mm were selected as the substrate material. The capillaries were rinsed with 18.2 MΩ·cm ultrapure water and then treated in a plasma cleaner for 30 minutes to thoroughly remove surface organic impurities and particulate matter. The cleaned capillaries were then fixed on a Sutter P-2000 laser drawing instrument. Utilizing the synergistic effect of laser heating and mechanical stretching, the four-hole structure was geometrically reduced to form a probe with a single-sided nano-tip. After drawing, the capillary tip formed a conical structure with a diameter of 80–300 nm. The original four micrometer-sized channels inside evolved into four independent and physically isolated nanochannels (pore sizes of approximately 20–60 nm) at the tip.

[0044] Next, carbon deposition is performed on the device using organic gas pyrolysis carbonization technology. The tip of the nanoprobe is placed in a protective gas (argon) environment, and butane gas (carbon source) is introduced into the channel. The outside of the capillary tip is locally heated using a butane flame until the tip turns bright yellow and is maintained for about 30 seconds. At high temperature, the butane gas undergoes pyrolysis on the inner wall of the nanochannel, depositing a dense conductive carbon layer. By precisely controlling the deposition time, four carbon nanoelectrodes are formed on the tip face of the channel, with the physical spacing between each pair of electrodes controlled within the range of 20-50 nm. After the device cools down, metal wires (such as 0.2 mm diameter copper wires) are inserted into the end of the channel to ensure close contact with the internal carbon electrodes, and then fixed and sealed with silicone or conductive adhesive. These wires serve as leads for subsequent electroplating processes, electrical detection, and dielectric release.

[0045] (2) Preparation of quantum tunneling detection electrode pairs (channel 1 and channel 3) The tip of a nanopipette is immersed in an electrolytic plating solution containing noble metal ions (such as gold ions). A reduction potential or constant current is applied via an electrochemical workstation, driving the metal ions in the plating solution to undergo a reduction reaction on the carbon electrode surface, depositing a metal nanolayer (such as gold nanoparticles or a gold film). During deposition, the growth of the metal layer on the carbon electrode surface gradually reduces the physical distance between the two electrodes. By monitoring changes in the deposition current in real time or employing a feedback control algorithm, the deposition time is precisely controlled. Ultimately, a carbon-supported gold nanotunneling electrode pair is formed at the tips of channels 1 and 3 of the four-hole capillary. The gap between the electrode pairs is reduced to 0.5–3 nm. This structure combines the mechanical stability of the tight bond between the carbon electrode and the glass tube wall with the chemical property of the gold surface, which easily anchors analyte molecules through sulfur-gold (Au-S) bonds, making it particularly suitable for constructing single-molecule conductive pathways. After preparation, the device is stored in ultrapure water for later use.

[0046] (3) Construction of dielectrophoretic enrichment units (channel 2 and channel 4) The remaining pair of symmetrical carbon-filled channels (channel 2 and channel 4) are used to construct a dielectrophoresis field. Since the openings of channels 2 and 4 are located at the nanotip, the applied alternating current generates a non-uniform electric field with an extremely high gradient in the tip region (i.e., dielectrophoresis tweezers). This electric field can overcome Brownian motion and actively capture and enrich biomolecules in complex matrices (such as urine) to the probe tip region.

[0047] (4) Prepare an aqueous solution of L-cysteine ​​with a concentration of 5-10 mM. Immerse the prepared four-hole quartz capillary tip in the solution and incubate at room temperature for 4-8 h. This allows cysteine ​​molecules to form stable Au-S covalent bonds with the gold layer on the surface of the carbon-based gold composite electrode of the tunneling detection unit through thiol (-SH) groups, thus constructing a dense anti-adsorption molecular layer on the surface of the tunneling electrode. After incubation, rinse the capillary tip repeatedly with ultrapure water to remove unbound free cysteine. Dry it with nitrogen gas for later use. This molecular layer can effectively block the non-specific adsorption of impurities such as proteins and cell debris in urine on the electrode surface, thus avoiding electrode contamination.

[0048] (5) Final form of the device and anti-interference design The final device is physically separated from the tunneling current detection circuit (channels 1 and 3) by a quartz wall (e.g., Figures 2-3 This achieves spatial insulation under the same body integration, greatly reducing the influence of strong AC electric field on the tunneling current detection circuit. The dielectric force located in channels 2 and 4 "drags" the target molecule to the enrichment area near the tip; the electrode pair located in channels 1 and 3 records its electrical conductivity fingerprint the instant the molecule flows through the tunneling gap.

[0049] (6) Characterization test The device obtained above was prepared using SEM. The nanoscale morphology of the four-hole structure at the tip was observed under SEM. Simultaneously, the device was placed under an optical microscope to observe the micron-scale macroscopic conical morphology and size of the tip. The results are as follows: Figure 4 As shown, the four-hole structure of the capillary tip is completely preserved, with no collapse, deformation or fusion of the channels; the tip has good sharpness, is uniformly conical, and has no defects such as bending or breakage.

[0050] Example 2: A method for measuring the conductivity of a non-targeted single molecule. This embodiment uses the non-targeted single-molecule conductivity spectroscopy measurement device prepared in Example 1 for detection. The detection flowchart is shown below. Figure 5 The specific process is as follows: (1) Sample pretreatment: Collect liquid samples containing target nanoparticles or biomolecules (including purification buffer system and complex biological fluids / matrices such as urine and serum), place them in a centrifuge, and centrifuge at 3000 rpm for 20 min; after centrifugation, take the supernatant as the sample to be tested. The original solution can be used directly, or diluted 5 times with 1×PBS buffer (pH=7.2~7.4) for later use.

[0051] (2) Single molecule capture and electrical signal acquisition: The prepared micro-nano electrical device is placed in the sample cell, and a DC bias voltage of 100 mV is applied to the tunneling electrode pair; a high-frequency AC voltage of 10 Vpp and 100 kHz is applied to the DEP electrode pair around the device, and dielectrophoretic capture is turned on. After 10~20 s, the AC power is turned off, and the electrical signal of the tunneling electrode pair is acquired. The acquisition time is 5 min. After completing one round of signal acquisition, the AC power is turned on again to complete a new round of signal acquisition. One sample to be tested is subjected to more than 4 rounds of capture and signal acquisition to obtain multiple sets of raw electrical signal data.

[0052] (3) Electrical signal algorithm processing For the acquired electrical signals, all information is merged, and wavelet transform is performed on the global signal sequence to remove baseline drift. Then, the sequence is continuously sliced ​​and multidimensional features are extracted using a constant-width sliding window, including time-domain statistical features, Fourier spectrum features, power spectrum features, wavelet features, and nonlinear features. Finally, classification is achieved through supervised machine learning.

[0053] In this embodiment, urine samples were collected from 40 healthy volunteers and 40 patients with interstitial cystitis, and randomly divided into a test group and a validation group at a ratio of 1:1. The test group included 20 healthy volunteers and 20 patients with interstitial cystitis, and the validation group included 20 healthy volunteers and 20 patients with interstitial cystitis.

[0054] Each test group sample underwent non-targeted single-molecule conductivity spectroscopy measurement using the method described above. The electrical signal spectra of two healthy volunteer samples (Sample 1 and Sample 3), two interstitial cystitis patient samples (Sample 2 and Sample 5), and one healthy sample (Sample 4) selected from the test group are shown below. Figures 6-10 As shown, the urine composition of healthy volunteers was stable, the baseline of nanopore detection was stable, the noise level was low, and there were no continuous high-amplitude pulse signals. The urine of patients with interstitial cystitis showed a significant increase in inflammatory components, and the detection showed large baseline fluctuations, high noise levels, and multiple high-amplitude signal bursts. Repeated verification of the signal characteristics of healthy volunteer samples proved the accuracy and stability of the detection method.

[0055] A machine learning model was built using Python software. The model parameters were: number of decision trees (n_estimators) = 100, maximum tree depth (max_depth) = 15, minimum number of samples required for internal node splitting (min_samples_split) = 5, and the model was a Random Forest classifier. In terms of specific feature collection, this model comprehensively characterizes the electrical response of target biomarkers (such as TNF-α and APF) in the complex urinary matrix as they pass through tunneling gaps. At the time-domain statistical feature level, it extracts peak counts, mean blocking amplitudes, and residence times by setting adaptive thresholds, and calculates the skewness and kurtosis of the global sequence to reflect the steric hindrance and spatial conformation of molecules. At the Fourier spectrum feature level, it extracts the 0–5 kHz dominant frequency band energy to distinguish characteristic tunneling events from slow collisions with polar impurities. At the power spectrum feature level, it calculates the power spectral density (PSD) to quantify charge transfer energy. At the wavelet feature level, it uses discrete wavelet transform (DWT) to extract high-frequency detail coefficients, capturing the transient waveforms of instantaneous folding / unfolding of molecules. At the nonlinear feature level, it extracts the Hjorth parameter to quantify temporal regularity and calculates the fractal dimension based on the Higuchi algorithm to measure self-similarity. For classification, this random forest model does not rely on a single electrical physical quantity but outputs a predicted probability score indicating whether a sample belongs to a pathological class. In this embodiment, the classification threshold is set to 0.5: when the model output probability threshold is greater than 0.5, the urine sample is judged to be a positive sample from a patient with interstitial cystitis; when it is lower than the threshold, it is judged to be a negative sample from a healthy population.

[0056] The detection results for different feature combinations are shown in Table 1 below: Table 1. Impact of different feature combinations on detection results According to the table above, when time-domain statistical features, Fourier spectrum features, power spectrum features, wavelet features, and nonlinear features are used together to construct a machine learning model, the model achieves the best results in sample classification, with the highest accuracy and consistency.

[0057] The machine learning model built in this example was used to test each sample in the test group according to the above method. After classification by the model, the accuracy reached 92% and the consistency reached 0.88. The sensitivity for judging patients with interstitial cystitis reached 93.3% and the specificity reached 90.0%. The sensitivity for judging healthy people reached 90.7% and the specificity reached 92%.

[0058] The results were validated in another batch of the same number of validation samples, with an accuracy of 90% and a consistency of 0.85. The sensitivity for diagnosing patients with interstitial cystitis reached 94% and the specificity reached 95.0%, while the sensitivity for diagnosing healthy individuals reached 96% and the specificity reached 93.0%.

[0059] Example 3: The Influence of Non-Targeted Single-Molecular Conductivity Spectroscopy Device on Detection This embodiment uses the method provided in Example 1 to prepare the apparatus, and conducts experiments according to the experimental method in Example 2. The apparatuses are divided into the following four groups based on their preparation methods: Group 1 (single tunneling): Traditional single tunneling electrode devices are used, and the preparation method is the same as in Example 1, except that the construction of dielectrophoretic enrichment units is not performed; Group 2 (Electrophoresis-tunneling): Electrophoresis-tunneling electrode devices are used. The preparation methods (1) and (2) are the same as in Example 1. Step (3) is to replace the carbon electrodes of channels 2 and 4 of the four-hole quartz capillary with electrophoresis-specific electrode pairs: platinum (Pt) microelectrodes are used, and two parallel electrodes are prepared on the tip end face of channels 2 and 4 by electroplating process. The spacing is consistent with the dielectric electrode spacing in Example 1. Group 3 (electrochemical tweezers-tunneling): Electrochemical tweezers-tunneling electrode device is used. The preparation methods (1) and (2) are the same as in Example 1. Step (3) is to prepare a special electrode pair for electrochemical tweezers in channels 2 and 4: a carbon-based gold composite electrode (the same material as the tunneling electrode pair) is used, but the electrode surface needs to be coated with a layer of polydopamine (PDA) to enhance molecular adsorption force; the electrode pair spacing is the same as the dielectric electrode spacing in Example 1.

[0060] Group 4 (dielectrophoresis-tunneling): using the apparatus provided in Example 1.

[0061] The experiment was conducted according to the experimental method in Example 2. Five healthy urine samples were selected, and each sample was monitored for 5 minutes under the same parameters. Five rounds of signal acquisition were performed on one sample, and the average value was taken. A machine learning model was constructed based on the detection spectrum. The acquisition results are shown in Table 2 below: Table 2. Influence of the conductivity spectrum measurement device on the dielectrophoresis-tunneling electrode As shown in Table 2, the target molecules in the urine samples of group 4 were rapidly enriched by dielectrophoresis, and the number of effective signals for capturing single molecules increased by more than 30 times; the capture response time was significantly shortened by about 20 times (from 250 s to 12.5 s), and the capture efficiency and signal-to-noise ratio were improved by 11.3 times and 5.0 times, respectively. Therefore, the active enrichment effect of dielectrophoresis in this invention solves the core technical bottleneck of traditional tunneling devices that "rely on random diffusion, have limited mass transfer, and have low capture efficiency".

[0062] Example 4: The Influence of Dielectrophoresis and Nanogap Electrodes on Different Spatial Arrangements This embodiment uses the measuring device provided in Embodiment 1 and conducts experiments according to the experimental method in Embodiment 2. When fabricating the dielectrophoresis-tunneling electrode device, the spatial arrangement of the dielectrophoresis and nano-gap electrode pairs is divided into the following three groups, as shown in the schematic diagram. Figure 11 : Group 1 (Orthogonal): Using the apparatus provided in Example 1; Group 2 (adjacent): During laser drawing of capillaries, four adjacent parallel channels are adjusted to be parallel to the capillary axis to ensure that the extension direction of the channels is consistent and without deviation after drawing. The tunneling electrode pair is set in channels 1 and 2, and the dielectrophoresis electrode pair is set in channels 3 and 4, so that the extension direction of the tunneling electrode pair and the dielectrophoresis electrode pair is completely consistent. The two electrode pairs are distributed side by side, and the two electrode pairs are isolated by the quartz tube wall. The gap between the electrode pairs is consistent with that of the experimental group. Group 3 (nested): The tunneling electrode pair (channels 1 and 3) serves as the main electrode, extending along the length of the capillary. During carbon deposition and electroplating, a gap of 5-10 nm is reserved in the tip electrode region. The dielectrophoretic electrode pair (channels 2 and 4) serves as the secondary electrode, with its tip embedded in the reserved gap of the tunneling electrode pair. The extension direction is perpendicular to the tunneling electrode pair, forming a nested structure where "you are in me". During laser drawing of the capillary, the channels of channels 2 and 4 bend from both sides (width direction) towards the center when they approach the tip, and the tip position is precisely aligned with the reserved gap of channels 1 and 3. The overlap area of ​​the two electrode pairs is controlled at the center of the capillary tip. Before carbon deposition, a plasma etching technique is used to grow an ultrathin quartz insulating layer on the inner wall of the reserved gap between channels 1 and 3 for isolation, to avoid short circuits, and to ensure the synergistic effect of the dielectrophoretic electric field and the tunneling detection electric field. The final electrode pair gaps are consistent with those of the experimental group.

[0063] Twenty urine samples were collected from healthy individuals and twenty from patients with interstitial cystitis. The experimental parameters for the three devices were completely identical. Each device was monitored for 5 minutes, and five rounds of signal acquisition were performed on each sample. The capture response time, capture efficiency, and stability were recorded, and the average value was taken. The experimental results are shown in Table 3 below: Table 3. Effects of different orientations of the dielectrophoresis-tunneling electrode As shown in Table 3, the orthogonal orientation (group 1) has the shortest capture response time of 12.5 s, and its capture efficiency and stability are significantly better than those of adjacent orientation and nested orientation. Furthermore, the orthogonal orientation also performs best in terms of the accuracy of sample classification.

[0064] The main reason is that the two electrode pairs in the orthogonal orientation extend perpendicularly, and the dielectrophoretic electric field is precisely aligned with the tunneling detection area. The non-uniform electric field gradient force generated by dielectrophoresis can directly drag the target molecules to the vicinity of the tunneling gap without the need for additional molecular diffusion, thus shortening the response time. However, the two electrode pairs in the adjacent orientation are distributed side by side, and there is a spatial gap between the dielectrophoretic enrichment area and the tunneling detection area. After the target molecules are enriched, they need to diffuse to the detection area through Brownian motion, which leads to a significant increase in response time. In addition, the electrode surface has a larger exposed area, and impurities in urine are easily adsorbed on the electrode surface. Long-term detection will lead to slight corrosion and affect the electrode life. The two electrode pairs in the nested orientation partially overlap. Although the enrichment area and the detection area are close, the electric fields in the overlapping area interfere with each other, weakening the capture efficiency of the dielectrophoretic force. This results in a longer response time than the orthogonal orientation. Moreover, the electric field strength in the electrode overlapping area is too high. Long-term application of the electric field will cause the insulation layer to break down and leakage to occur. At the same time, impurities are easy to accumulate in the overlapping area and corrode the electrode.

[0065] Example 5: The effect of the number of dielectrophoretic electrodes on detection This embodiment uses the measuring device provided in Embodiment 1 and conducts experiments according to the experimental method in Embodiment 2. The dielectrophoresis electrodes are divided into three groups: 1, 2, and 4. (With 1 electrode, channel 2 or channel 4 is selected as the preparation channel for a single dielectrophoresis auxiliary electrode, and the remaining channels are insulated and sealed to avoid interference with the electric field distribution from the solution. With 4 electrodes, channels 2 / 4 are each divided into dual-electrode channels. Two parallel sub-channels are prepared within the original channel 2 / 4 through micro / nano etching, ultimately forming 4 independent dielectrophoresis auxiliary electrode channels. See the schematic diagram below.) Figure 12 Twenty urine samples were collected from healthy individuals and twenty from patients with interstitial cystitis. All experimental parameters were identical. The capture response time, capture efficiency, and accuracy were recorded. The experimental results are shown in Table 4 below. Table 4. Effect of the number of dielectrophoretic electrodes on detection. As shown in Table 4, the electric field generated by a single electrode is non-uniform, with the electric field strength being maximum only near the electrode surface and rapidly decaying away from the electrode. It cannot form a stable gradient field, resulting in a very small electric field coverage area compared to multi-electrode systems. This only captures target molecules near the electrode surface, leading to a significantly longer response time, reduced efficiency, and susceptibility to interference from impurities in the solution, thus lowering accuracy. Compared to four electrodes, two electrodes (a pair of dielectrophoresis auxiliary electrodes) can form a perfect orthogonal symmetric structure with the main tunneling electrode, creating a highly concentrated and strongest electric field gradient at the very center of the tunneling gap. This allows target proteins in urine to be precisely and efficiently focused on the nanoscale detection area, achieving the optimal single-molecule capture efficiency (92.5%) and extremely high classification accuracy (95.0%). While theoretically the spatial electric field is stronger when the number of electrodes increases to four (response time slightly shortened to 10.0 s), the excessively dense quadrupole electric field distribution causes severe distortion of the local electric field. This complex overlapping electric field not only easily induces strong Joule heating at the extremely small capillary tip, disrupting the natural conformation of biomarkers, but also leads to the over-enrichment of a large number of urinary background impurity proteins in non-sensing regions (i.e., creating a dead zone effect). This increases background noise and steric hindrance, causing the effective capture efficiency into the tunneling gap to drop to 85.0%, and the accuracy also declines accordingly. Therefore, considering both the difficulty of micro / nano fabrication and actual sensing performance, setting up two dielectrophoresis auxiliary electrodes is the optimal configuration for achieving precise capture of target molecules in complex liquid matrices.

[0066] Furthermore, by adjusting only the gap between the tunneling electrodes and setting up two dielectrophoresis electrodes while keeping other parameters constant, it was found that when the gap was less than 0.5 nm, the tunneling performance was too poor to be usable, while when the gap was greater than 3.5 nm, there was almost no current detected. This is because when the gap is too small, the steric hindrance prevents the dielectrophoretically enriched target molecules from entering the gap to form a bridge, directly leading to capture failure and a low capture success rate. When the gap is too large, the quantum tunneling effect is significantly weakened, and no recognizable characteristic conductance signal can be generated when single molecules bridge, resulting in a significant decrease in detection sensitivity and even an inability to distinguish between background noise and molecular signals, leading to an extremely low effective signal recognition rate. Therefore, an electrode gap of 0.5~3 nm is the optimal choice.

[0067] Example 6: Single-molecule detection of different samples This embodiment uses the measuring device provided in Embodiment 1 and conducts the experiment according to the experimental method in Embodiment 2. The urine samples are divided into the following five treatment groups: Group 1: Raw urine sample; Group 2: High-impurity urine samples, containing a large amount of cell debris and protein precipitates; Group 3: Low-concentration urine samples, with biomarker concentrations of approximately 0.1 nM, purified by ultrafiltration; Group 4: 5 nm radius gold nanoparticles (AuNPs), concentration gradient up to 10 fM (10 femtomoles / L). Group 5: Human serum albumin (HSA), concentration gradient up to 33 pM (33 pmol / L) The acquisition maps of the AuNPs group and the HSA group are as follows: Figures 13-14 As shown, without DEP, extremely low concentration targets exhibit only baseline noise and no effective detection signal due to limited random diffusion mass transfer. After enabling DEP, the current signal is significantly enhanced, exhibiting a dense and stable characteristic response, and the effective signal density / recognition rate is greatly improved. The experimental results are shown in Table 5 below: Table 5 Single-molecule detection in different urine samples As shown in Table 5, the device and detection method of this invention can achieve efficient and stable capture and detection of target single molecules under conditions such as extremely low concentration (0.1 nM) targets, high impurity samples, and original biological matrix samples. The capture success rate is greater than 80%, and the signal-to-noise ratio is maintained at an extremely high level of over 12.0. Furthermore, it significantly simplifies the sample pretreatment process and shortens the detection time. This indicates that the present invention can break through the detection concentration limit of traditional tunneling devices at the nanometer level and above, achieve highly sensitive detection of extremely low concentration samples, and at the same time have excellent anti-impurity interference ability and are compatible with high impurity biological matrices, providing key experimental evidence for trace target detection and early clinical diagnosis.

[0068] Example 7: The Influence of Tunneling Electrode Surface Modification Strategy on Detection Results This embodiment uses the measuring device provided in Embodiment 1 and performs testing according to the experimental method in Embodiment 2. The following three groups are set up for the surface modification strategy of the tunneling electrode: Group 1: The tunneling electrode has no surface modification; Group 2: The tunneling electrode was modified with cysteine ​​using step (4) of Example 1; Group 3: The tunneling electrode is modified with mercaptoethanol.

[0069] Twenty high-impurity urine samples (same as Group 2 in Example 6, containing a large amount of cell debris and protein precipitation) were selected. The experimental parameters were completely consistent with those in Example 2. Each sample was collected for 5 minutes, and the average value was taken. The capture efficiency, signal-to-noise ratio, and signal drift rate were recorded and averaged. The experimental results are shown in Table 6 below: Table 6. Impact of Tunnel Electrode Surface Modification Strategies on Detection As shown in Table 6, the unmodified electrode in Group 1 is prone to non-specific adsorption in urine with high impurities. With the increase of detection rounds, the electrode surface is covered by cell debris and protein, resulting in a significant decrease in capture efficiency and signal-to-noise ratio, high signal drift rate, and inability to achieve continuous and stable detection. Since the strong alternating electric field accelerates the directional migration of target molecules and impurities to the tunneling detection region (nano-gap electrode pair), the surface of the tunneling electrode is exposed to a higher concentration of molecular flux, which significantly aggravates non-specific adsorption. Therefore, targeted anti-adsorption modification of the tunneling electrode is a necessary prerequisite for ensuring the stability of detection in complex matrices.

[0070] Groups 2 and 3 were both treated with anti-adsorption, but different modification strategies had different effects on the dielectrophoresis system. As can be seen from the table, although mercaptoethanol modification can bind to the gold layer surface of carbon-based gold composite electrode through Au-S bond and has a certain anti-adsorption effect, it will also shield and distort the dielectrophoresis electric field, reduce the capture efficiency, and its stability is poor. It is easy to fall off the electrode surface after incubation. As the number of detection rounds increases, the anti-adsorption effect gradually weakens. On the other hand, the cysteine-modified group has the highest capture efficiency, signal-to-noise ratio and lowest signal drift rate. It can effectively suppress the adsorption of tunneling electrode without interfering with the electric field generation and molecular enrichment function of dielectrophoresis electrode. Moreover, in terms of sample classification, the classification accuracy is the highest when the electrode surface is modified with cysteine.

[0071] In summary, modifying the surface of the tunneling electrode with cysteine ​​yields the best results.

[0072] Example 8: Dielectrophoretic Manipulation-Nano-Gap Tunneling Detection Conductivity Spectroscopy Device Based on Planar Chip This embodiment, as another implementation of the present invention, retains the core coupling mechanism of "dielectrophoresis (DEP) manipulation + nano-gap tunneling detection" of the present invention. Unlike the three-dimensional structure of the four-hole quartz capillary in Embodiment 1, it employs planar micro-nano fabrication technology to fabricate the device on a planar insulating substrate, enabling mass production, integration, and automated detection of the device. This is suitable for clinical batch testing scenarios of urine samples. The specific implementation steps are as follows: (1) Substrate pretreatment Take a silicon wafer substrate and ultrasonically clean it for 10 minutes each with acetone, anhydrous ethanol, and deionized water to remove surface oil, impurities, and oxide layers. Dry the substrate surface with nitrogen and place it in a plasma cleaner with a power of 100W for 5 minutes to activate the substrate surface and improve the adhesion of subsequent photoresist and electrode materials. Place the pretreated substrate in an oven and bake it at 80°C for 30 minutes to remove residual moisture and set it aside.

[0073] (2) Fabrication of tunneling detection unit (planar nanoelectrode pair) A layer of positive photoresist was uniformly coated onto the pretreated silicon substrate at 3000 rpm to a thickness of 1 μm. The substrate was then preheated in an oven at 90°C for 60 s to ensure uniform curing of the photoresist. The substrate was then placed in an electron beam lithography system, and the tunneling electrode pair pattern was applied. Parameters such as exposure dose and acceleration voltage were set, with a focus on controlling the electrode gap size to be 0.5–3 nm to complete the nanoscale pattern exposure. The exposed substrate was then placed in a photoresist developer for 60 s to remove unexposed photoresist, exposing the area to be deposited as electrodes. After rinsing with deionized water, the substrate was preheated in an oven at 100°C for 30 min to cure the photoresist pattern. The substrate was then placed in a magnetron sputtering system, where a 5 nm chromium layer (adhesion layer) was deposited first, followed by a 50 nm gold layer to ensure good electrode conductivity. The substrate was then placed in a stripping solution and ultrasonically stripped for 10 min to remove residual photoresist, yielding a planar nano-tunneling electrode pair. The gap between the tunneling electrodes was finely modified using a focused ion beam (FIB) to correct any gap size deviations. SEM characterization confirmed that the electrode structure was intact, the gap size was uniform, and it met the design requirements.

[0074] (3) Fabrication of dielectrophoretic manipulation units (microelectrode array) On the substrate with the prepared tunneling electrode pair, a positive photoresist was coated again at a rotation speed of 2000 rpm and a coating thickness of 2 μm. The substrate was then pre-baked at 90°C for 60 s. The substrate was placed in a photolithography apparatus and a microelectrode array pattern (interdigitated microelectrode array, distributed around the tunneling electrode pair, with an electrode width of 10 μm and a spacing of 5 μm) was loaded to complete the micron-level pattern exposure. Step (2) was repeated to deposit a chromium-gold electrode layer and remove the photoresist to obtain an interdigitated DEP microelectrode array integrated on the same planar substrate. The conductivity of the DEP microelectrode array was tested using an electrochemical workstation to ensure that the electrodes were free from open circuits and short circuits and could stably generate a non-uniform alternating electric field.

[0075] (4) Anti-adsorption modification Prepare a 10 mM cysteine ​​PBS buffer (pH=7.2~7.4), immerse the silicon substrate of the prepared DEP microelectrode array in the buffer, and incubate at 37℃ for 4~8 h to allow cysteine ​​to form Au-S covalent bonds with the carbon-based gold composite electrode surface through thiol groups, forming a uniform anti-adsorption modification layer on the tunneling electrode surface. After incubation, rinse the substrate three times with PBS buffer, then rinse with deionized water, and dry with nitrogen. Test the electrochemical impedance of the modified electrode using an electrochemical workstation to confirm the compactness of the modification layer and ensure that there is no bare electrode surface.

[0076] (5) Fluid channel fabrication and integration SU-8 photoresist was coated on a silicon wafer, and after photolithography exposure and development, a microfluidic channel mold was prepared (channel width 100 μm, depth 50 μm, length 1 cm, covering the tunneling electrode pair and DEP microelectrode array area); PDMS prepolymer and curing agent were mixed uniformly at a volume ratio of 10:1 and poured onto the mold. After vacuum degassing for 10 min, it was placed in an oven at 80℃ for 2 h for curing; the cured PDMS was peeled off from the mold, and the microfluidic channels were cut out. The surface of the PDMS channel and the surface of the device substrate were activated by a plasma cleaner (power 100W, treatment for 3 min); the PDMS microfluidic channels were precisely aligned with the substrate (ensuring that the channels cover the tunneling electrode pair and DEP microelectrode array), pressed and bonded, and placed at room temperature for 30 min to achieve sealed integration of the fluid channel and the planar device, resulting in a complete planar micro / nano electrical device.

[0077] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A non-targeted single-molecule conductivity spectroscopy measurement device, characterized in that, It includes a single-molecule enrichment unit and a tunneling detection unit; the single-molecule enrichment unit is used to capture, enrich, or locate single molecules in the sample to be tested to or near the nano-gap; the tunneling detection unit is used to perform non-targeted detection of single molecules entering the nano-gap; the single-molecule enrichment unit has nanoscale control precision, including any one or more of dielectrophoresis, electrophoresis, and electrochemical tweezers.

2. The measuring device as described in claim 1, characterized in that, The single-molecule enrichment unit is a dielectrophoretic manipulation unit, including 1 to 4 auxiliary electrodes; the tunneling detection unit includes a pair of nano-gap electrodes; the gap width of the nano-gap electrode pair is sub-5 nanometers, and its surface is modified with an anti-adsorption molecular layer of cysteine ​​or mercaptoethanol; the auxiliary electrodes and the nano-gap electrode pair are arranged orthogonally, nested or adjacent to each other in space.

3. The measuring device as described in claim 2, characterized in that, The device further includes an insulating substrate, through which the dielectric electrophoresis manipulation unit and the tunneling detection unit are insulated from each other; the insulating substrate is made of any one or more of quartz, glass tube, silicon wafer, glass sheet, polymer, and flexible material.

4. A device for measuring non-targeted single-molecule conductivity spectra, characterized in that, It includes the measuring device, microfluidic channel sample cell and support as described in any one of claims 1 to 3, and the electrical control module.

5. The measuring device as described in claim 4, characterized in that, The microfluidic channel sample cell includes a glass tube with an inner diameter of 1.5 mm to 2.5 mm and its support frame; the support frame tilts the glass tube at an angle of 10 to 30 degrees; it also includes an electrical control module, which includes a working electrode, a reference electrode, a function signal generator, a patch-clamp amplifier, and a digital-to-analog converter.

6. Use of the measuring device according to any one of claims 1 to 3 or the measuring apparatus according to any one of claims 4 to 5 in non-targeted detection of complex matrices.

7. A method for measuring non-targeted single-molecule conductivity spectra, characterized in that, The measurement device as described in any one of claims 1 to 3 or the measurement apparatus as described in any one of claims 4 to 5 is used for detection. The sample to be tested is brought into contact with the measurement device in a microfluidic channel sample cell to obtain conductivity spectroscopy characterization, thereby achieving non-targeted characterization of single-molecule in-situ, real-time electronics in the sample to be tested.

8. The measurement method as described in claim 7, characterized in that, The method includes the following steps: (1) The sample to be tested and the measuring device are brought into contact in the microfluidic channel sample cell, the tunneling detection unit is kept energized, the dielectrophoresis manipulation unit is turned on to apply an alternating electric field, and the molecules in the sample are driven and enriched into the detection gap. (2) Turn off the dielectrophoresis manipulation unit, and obtain the non-targeted single-molecule conductivity spectrum of the sample to be tested through the tunneling detection unit. Repeat steps (1) and (2) more than 4 times to maintain the molecular capture probability in the detection gap and accumulate single-molecule conductivity spectrum data. (3) After analysis and feature extraction, a model is constructed to classify the samples to be tested.

9. The measurement method as described in claim 8, characterized in that, The peak-to-peak value of the sinusoidal AC signal of the dielectrophoresis manipulation unit is 1~20 Vpp, and the frequency is 100 kHz~5 MHz; the time for applying the AC electric field in step (1) is 5~20 s; the bias voltage applied to the tunneling detection unit in step (2) is 50~300 mV, and the collection time for obtaining the non-targeted single-molecule conductivity spectrum of the sample to be tested is 4~10 min.

10. The measurement method as described in claim 9, characterized in that, The features include feature parameters extracted from multiple dimensions such as time-domain statistical features, Fourier spectrum features, power spectrum features, wavelet features, and nonlinear features; the test sample is any one or more of urine, serum, saliva, milk, or wine; the classification of the test sample includes any one or more of the following: differentiation between disease and health, differentiation between different types of disease, pathogen identification, and cancer subtyping.

Images