Laboratory particle-gas cooperative detection method based on aerosol adsorption mechanism

Abstract

The invention discloses a laboratory particle-gas cooperative detection method based on an aerosol adsorption mechanism, and relates to the technical field of gas detection. The method comprises the following steps: step 1, conveying a to-be-detected sample airflow at a constant speed, enabling the to-be-detected sample airflow to pass through a blank integrated adsorption carrier, and collecting a pure background signal; 2, a layered hydrophilic-hydrophobic adsorption film is deposited on the surface of the integrated adsorption carrier in situ, and the functionalized integrated adsorption carrier for laboratory particle-gas collaborative detection is obtained; step 3, enabling the to-be-tested sample airflow to pass through the functionalized integrated adsorption carrier in a pulse mode; the method comprises the following steps: obtaining an original particle-gas synergistic signal through optical-acoustic double-path excitation by utilizing a micron-sized spiral pore channel of a layered hydrophilic-hydrophobic adsorption film; and 4, performing multi-section spiral vector trade-off processing on the obtained original particle-gas cooperative signal. The system is quick in response, high in interference resistance and suitable for high-reliability particle-gas cooperative monitoring in a complex environment of a laboratory.

Description

Technical Field

[0001] The present invention relates to the technical field of gas detection, and in particular to a laboratory particle-gas collaborative detection method based on an aerosol adsorption mechanism. Background Art

[0002] In the field of laboratory environmental safety testing, the simultaneous monitoring of particulate pollutants and volatile organic gases has become an important technical requirement, especially in scenarios such as microelectronics, biopharmaceuticals, and fine chemicals that have extremely high requirements for air cleanliness and component stability. Traditional particle detection technologies are mostly based on the principle of light scattering, inertial impact sampling, or mass gravity weighing. They have strong particle size response capabilities and can achieve concentration monitoring and statistical analysis of submicron to micron solid particles. However, these technologies can often only perceive particle phase information independently and cannot identify chemical gas components that are simultaneously present in the airflow. In particular, under conditions of low concentrations and strong reactivity of volatile organic compounds (VOCs), sulfides, amine gases, etc., it is difficult to form a comparable collaborative data model.

[0003] On the other hand, detection methods for gas components focus more on chemical selectivity and sensitivity. Mainstream technologies include gas chromatography, mass spectrometry, electrochemical sensors, metal oxide semiconductor sensors, and spectral absorption devices. These methods generally rely on separation and enrichment or chemical reactions. Although they perform well in terms of gas detection sensitivity, they have long response times, are sensitive to changes in temperature and humidity, and lack the ability to output structured signals, making them unable to meet the needs of simultaneous dynamic monitoring of multiple components. In addition, most existing gas detection methods do not consider the cross-influence issues when gases and particles coexist. In real laboratory environments, particles often exist as gas adsorption or transmission carriers, and there is a complex interaction between the two. Collecting gas signals alone and ignoring particle behavior will seriously limit the safety of system operation and the ability to track pollution sources. Summary of the Invention

[0004] In view of this, the present invention provides a laboratory particle-gas collaborative detection method based on the aerosol adsorption mechanism. By in situ depositing a layered amphoteric-philic adsorption film on the surface of an inert porous substrate and inducing the formation of micron-scale spiral channels, the simultaneous enrichment and unified analysis of submicron particles and volatile gas components are achieved. Combining optical-acoustic dual-path excitation with multi-segment spiral vector compromise processing, the system can map complex collaborative signals into discriminable characteristic maps, significantly improving detection sensitivity, data consistency and anomaly recognition capabilities. This method has a compact structure, fast response, and strong anti-interference ability, and is suitable for high-reliability particle-gas collaborative monitoring in complex laboratory environments.

[0005] The technical solution adopted in the present invention is as follows:

[0006] A laboratory particle-gas collaborative detection method based on an aerosol adsorption mechanism, the method comprising:

[0007] Step 1: The sample airflow to be tested is transported at a constant speed through a blank integrated adsorption carrier to collect a pure background signal. The baseline of the particle-gas synergistic signal in the laboratory environment is established by time averaging and noise suppression of the pure background signal.

[0008] Step 2: In situ depositing a layered adsorption film with both affinity and repellency on the surface of the integrated adsorption carrier to form micron-scale spiral channels in the adsorption film; encapsulating the adsorption film with a multi-stage impact collection structure to obtain a functionalized integrated adsorption carrier for laboratory particle-gas collaborative detection;

[0009] Step 3: The sample gas flow to be tested is passed through the functionalized integrated adsorption carrier in a pulsed mode. The micron-scale spiral pores of the layered amphoteric-philic adsorption membrane are used to achieve simultaneous enrichment of submicron particles and volatile gas components. After the simultaneous enrichment is completed, the original particle-gas synergistic signal is obtained through optical-acoustic dual-path excitation.

[0010] Step 4: Perform multi-segment spiral vector compromise processing on the acquired original particle-gas synergistic signal, and map the final heterogeneous reconstruction matrix into a particle-gas synergistic feature map; compare the particle-gas synergistic feature map with the baseline one by one; if any feature exceeds the baseline threshold, the alarm logic is automatically triggered and a complete detection report is generated; if all features are within the baseline threshold, a safety status report is output and archived.

[0011] Furthermore, in step 1, time averaging and noise suppression processing are performed on the pure background signal obtained in the entire time range; the time averaging and noise suppression processing uses an equal-weight sliding window method to fully smooth the pure background signal according to a one-second time window, and linear interpolation is performed on the random spikes in each time window; after the processing is completed, a smoothed pure background signal is obtained; the maximum value of the smoothed pure background signal in the entire time range is used as the upper threshold, and the minimum value of the smoothed pure background signal in the entire time range is used as the lower threshold to construct the particle-gas synergistic signal baseline of the laboratory environment.

[0012] Furthermore, the blank integrated adsorption carrier is cut and formed as a whole from an inert porous substrate, which does not contain any surface functional groups, has a continuous pore structure that remains through and has no directional restrictions; when the blank integrated adsorption carrier is installed on an inert bracket, it only allows the sample airflow to be tested to enter from the calibrated air inlet surface and be discharged from the opposite single air outlet surface.

[0013] Furthermore, in step 2, a layered amphoteric-philic adsorption film is in situ deposited on the surface of the integrated adsorption carrier to form a micron-scale spiral channel in the adsorption film, which includes: preparing an amphoteric precursor solution and a phobic precursor solution by a proportioning solution method, and both precursor solutions are fully homogenized under constant temperature magnetic stirring without introducing any surfactant; then the amphoteric precursor solution and the phobic precursor solution are sequentially divided into light-proof containers as the only precursor source for in situ deposition; performing in situ deposition, including: depositing the amphoteric precursor solution, the phobic precursor solution and the step-by-step deposition on the inert porous substrate. Segmented air drying cycle; each completion of the affinity-phobic bilayer deposition is counted as one cycle, and the total number of cycles is not less than 5 times; the entire in-situ deposition process is carried out on a constant rotating platform, and the rotating platform maintains a constant angular velocity of 180 degrees per second at a uniform speed to ensure that the layered affinity-phobic amphoteric adsorption film is evenly covered on the entire surface and has consistent thickness; after the in-situ deposition is completed, the rotating magnetic field induction is triggered, and the rotating magnetic field runs in a clockwise direction for a duration of not less than 30 seconds; during the rotating magnetic field induction process, a directional rearrangement occurs inside the layered affinity-phobic amphoteric adsorption film, generating micron-scale spiral channels that penetrate the thickness of the adsorption film.

[0014] Furthermore, in step 2, the adsorption membrane and the multi-stage impact collection structure are integrally packaged to obtain a functional integrated adsorption carrier for laboratory particle-gas collaborative detection. The process includes: after the rotating magnetic field is released, it is allowed to stand for three minutes to allow the micron-scale spiral channel morphology to spontaneously stabilize; the layered amphoteric and amphoteric adsorption membrane forming micron-scale spiral channels is placed together with the inert porous substrate into a vacuum thermal curing cavity; under a negative pressure state, the temperature is gradually raised to 120 degrees Celsius and maintained for 45 minutes to allow the amphoteric and amphoteric interface to undergo cross-linking and curing, and after the cross-linking and curing is completed, the temperature is lowered to room temperature, and the cavity is maintained at a negative pressure during this period to prevent external water vapor from entering; the cross-linked and cured layered amphoteric and amphoteric adsorption membrane and the prefabricated multi-stage impact collection structure are integrally packaged by an isobaric interlocking method.

[0015] Furthermore, during packaging, the air inlet surface of the multi-stage impact collection structure is ensured to be parallel to the outer surface of the layered amphoteric-philic adsorption membrane, and there is no air gap between the two; an inert sealing gasket is used to achieve circumferential sealing to avoid airflow bypass.

[0016] Furthermore, in step 4, a multi-segment spiral vector compromise processing is performed on the obtained original particle-gas synergistic signal, and the process of mapping the final heterogeneous reconstruction matrix into a particle-gas synergistic characteristic map includes: dividing the original particle-gas synergistic signal into equal-length segments according to the pulse sequence, generating a spiral vector index for each segment, and uniquely identifying it with a segment number; performing cross-rank sorting on the spiral vector index of each segment to obtain a monotonically increasing segment priority list; comparing the baseline in the order of the segment priority list, performing progressive threshold gate update on each segment to form a heterogeneous reconstruction matrix; based on the progressive threshold gate update result, calling the compromise fusion rule, bidirectionally trimming the heterogeneous reconstruction matrix according to the priority, and obtaining the final heterogeneous reconstruction matrix; mapping the final heterogeneous reconstruction matrix into a particle-gas synergistic characteristic map.

[0017] Furthermore, an initial reference segment is arbitrarily selected from all segments, and the spiral vector index of the initial reference segment is placed as the reference index at the starting position of the segment priority list; for each of the remaining segments, the insertion rank position between its spiral vector index and the spiral vector index of each segment in the current segment priority list is calculated respectively, wherein the insertion rank position determination process includes: calculating the vector space Euclidean distance between the current segment spiral vector index and the spiral vector index of each segment in the segment priority list; determining the relative sorting position between the spiral vector index of the segment to be sorted and the two adjacent segments in the segment priority list according to the size of the Euclidean distance, ensuring that all insertion rank position calculation results remain monotonic and there is no cross-contradiction; repeating the above operation for all segments until the clear insertion rank positions of all segment spiral vector indices relative to the existing segments in the segment priority list are determined; according to all the clear insertion rank positions obtained, all segment spiral vector indices are inserted into the corresponding positions of the segment priority list in sequence to obtain a complete monotonically increasing segment priority list.

[0018] Furthermore, monotonically increasing is defined as follows: the Euclidean distances between the spiral vector indices of any two adjacent segments in the segment priority list all show a trend of gradually increasing.

[0019] By adopting the above technical solution, the present invention produces the following beneficial effects: First, by in-situ depositing a layered amphoteric-philic adsorption film on an inert porous substrate and inducing the formation of micron-scale spiral channels, the synchronous enrichment of submicron particles and volatile gas components is achieved in a single physical structure. This structure has both fluid stability and adsorption selectivity, significantly improving the synergistic enrichment efficiency. Secondly, the optical-acoustic dual-path excitation mechanism can trigger the synchronous release of particles and gases without destroying the channel structure, and form original synergistic signals with consistent timing and complementary responses, providing a unified data basis for signal analysis in time and space. Thirdly, through the multi-segment spiral vector compromise processing and heterogeneous reconstruction matrix construction mechanism, the complex original signal is mapped into a particle-gas synergistic characteristic map, solving the problem that heterogeneous data is difficult to express in a unified way, and improving the automatic discrimination and risk identification capabilities. The system uses baseline construction and progressive threshold gate update strategies, which have strong noise resistance and adaptability in dynamic environments, effectively reducing the false alarm rate and missed alarm risk. The overall structure utilizes inert materials, isobaric packaging, and a gap-free design, ensuring consistent and airtight detection paths. It also boasts strong mechanical stability and thermal reliability, making it suitable for long-term laboratory environmental monitoring scenarios. This invention integrates innovations in microscale structural design, synchronous enrichment mechanisms, collaborative signal analysis, and automatic alarm logic, providing a fast-response, clear-criteria, compact, and low-maintenance particle-gas collaborative detection solution for laboratory environmental safety management. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 1 is a schematic diagram of a method flow of a laboratory particle-gas collaborative detection method based on an aerosol adsorption mechanism in an embodiment of the present invention;

[0021] Figure 2 Schematic diagram of the three-dimensional structure of the micron-scale spiral channel inside the integrated adsorption carrier in an embodiment of the present invention;

[0022] Figure 3 Schematic diagram of the simultaneous enrichment process of submicron particles and volatile gas components in an embodiment of the present invention;

[0023] Figure 4 3 is a comparative analysis diagram of the particle-gas synergistic characteristic spectrum and the baseline threshold in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] All features disclosed in this specification, or all steps in the disclosed methods or processes, except mutually exclusive features and / or steps, can be combined in any manner.

[0025] Any feature disclosed in this specification (including any appended claims and abstract), unless otherwise stated, may be replaced by other equivalent or similar features. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

[0026] refer to Figure 1 :A laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism, the method comprising:

[0027] Step 1: The sample airflow to be tested is transported at a constant speed through a blank integrated adsorption carrier to collect a pure background signal. The baseline of the particle-gas synergistic signal in the laboratory environment is established by time averaging and noise suppression of the pure background signal.

[0028] Step 2: In situ depositing a layered adsorption film with both affinity and repellency on the surface of the integrated adsorption carrier to form micron-scale spiral channels in the adsorption film; encapsulating the adsorption film with a multi-stage impact collection structure to obtain a functionalized integrated adsorption carrier for laboratory particle-gas collaborative detection;

[0029] Step 3: The sample gas flow to be tested is passed through the functionalized integrated adsorption carrier in a pulsed mode. The micron-scale spiral pores of the layered amphoteric-philic adsorption membrane are used to achieve simultaneous enrichment of submicron particles and volatile gas components. After the simultaneous enrichment is completed, the original particle-gas synergistic signal is obtained through optical-acoustic dual-path excitation.

[0030] Step 4: Perform multi-segment spiral vector compromise processing on the acquired original particle-gas synergistic signal, and map the final heterogeneous reconstruction matrix into a particle-gas synergistic feature map; compare the particle-gas synergistic feature map with the baseline one by one; if any feature exceeds the baseline threshold, the alarm logic is automatically triggered and a complete detection report is generated; if all features are within the baseline threshold, a safety status report is output and archived.

[0031] The transport, interaction, and transient release of aerosol particles and volatile gas molecules are simultaneously manipulated within the same spatial and temporal framework, resulting in resolvable and complementary composite signals from both species at a single detection node. A layered, amphoteric-philic adsorption membrane plays the physical and chemical role in this coupling process: the amphoteric layer, due to its numerous polar sites, produces reversible hydrogen bonds and dipole-dipole adsorption for polar and semi-polar volatile molecules, while the amphoteric layer captures hydrophobic or low-polarity gases with a weak van der Waals potential, while exhibiting filter-like mechanical interception of submicron particles. When these two layers with significantly different properties are repeatedly deposited and induced by a rotating magnetic field to generate micron-scale spiral pores throughout the membrane thickness, a spiral-shear composite flow field forms within the pores, promoting lateral migration and inertial attachment of particles along the pore walls, while gas molecules are progressively enriched in a permeation-desorption-repermeation cycle. Because the hydrophilic interface has different surface energies, the potential barrier heights of particles and gases near the interface are different, resulting in a time delay distribution under pulse mode driving: the particle enrichment moment is earlier, and the gas enrichment moment is later. This time dislocation creates a distinguishable window for the signal during subsequent optical-acoustic dual-path excitation.

[0032] The blank integrated adsorption carrier provides an interference-free continuous pore structure, maintaining laminar aerosol flow and minimizing shear turbulence before delivery to the adsorption membrane. The collected pure background signal, after time averaging and noise suppression, forms a particle-gas synergistic signal baseline. This baseline not only reflects the normal coexistence spectrum of particles and gases in the laboratory environment but also records the low-frequency and high-frequency boundaries of the system's background noise. The significance of establishing a baseline is that any subsequent enrichment-analysis events must be quantified relative to this reference surface; otherwise, spurious offsets introduced by minor fluctuations in operating conditions will be misinterpreted as anomalies.

[0033] When the functionalized integrated adsorption carrier is encapsulated and switched to pulse mode, the airflow pulse first undergoes inertial classification on the multi-stage impaction collection structure. Large particle components are dissipated at the early stages to protect the integrity of the downstream spiral channel. Subsequent submicron particles and target gas molecules enter the spiral channel. Under pulsed flow with a Reynolds number close to the transition zone, the spiral geometry generates a superposition field of Coriolis force, centrifugal force, and shear force, causing the particle trajectory to spiral along the wall and be captured at the affinity-repellent interface. At the same time, gas molecules at the central axis of the channel, driven by the concentration gradient, continue to diffuse toward the interface and are temporarily adsorbed by the adjacent affinity layer. The pressure drop during the pulse interval promotes the desorption of some adsorbed gas and its redistribution in the channel, forming an "adsorption-desorption" oscillation and deepening the concentration enrichment ratio.

[0034] During the optical-acoustic dual-path excitation stage, a short pulsed light beam applies localized transient heating to the adsorption film. The difference in interfacial thermal expansion coefficients between the affinic and phobic sites generates microstress, prompting rapid desorption of the accumulated gas molecules. Simultaneously, the external acoustic field creates standing waves within the channel. Alternating pressure peaks and troughs periodically perturb the particle-gas mixture, further stripping away any remaining gas. The particles then microvibrate under the acoustic pressure gradient, collectively releasing the original particle-gas synergistic signal. This signal exhibits multi-peak pulses on the time axis, each corresponding to an accumulation-excitation-release cycle. The peak width is determined by the optical pulse width, the acoustic field frequency, and the length of the spiral channel.

[0035] To separate true anomalies from systematic drift from multi-peak signals, the algorithm employs a multi-segment spiral vector compromise process. First, the complete time series is divided into segments of equal length according to the pulse period. For each segment, multidimensional features such as amplitude, rising edge steepness, and half-width (FWHM) are extracted and encoded as spiral vector indices. Each index is essentially a projection point of a signal segment into vector space. The interpolation rank sorting process is equivalent to finding the sequence of points that most closely approximates a monotonically increasing path in high-dimensional space. This ensures that subsequent threshold updates avoid false detections caused by priority inversion between adjacent segments. The threshold update maps the upper and lower thresholds of the baseline to the current upper and lower floating intervals of each segment. The segment spiral vector index is then compared with this floating interval. If the index jumps outside the interval, it indicates that the segment contains an anomalous release. After progressive threshold gate updates, all segments are aggregated into a heterogeneous reconstruction matrix. Bidirectional clipping is performed using a compromise fusion rule to remove isolated but small-amplitude noise entries in different dimensions, retaining anomaly vectors with cross-dimensional consistency, and ultimately mapping the particle-gas synergistic characteristic map.

[0036] This characteristic spectrum simultaneously displays the changing trends of the segment number and the spiral vector index in a two-dimensional coordinate system. If the spectrum curve breaks through the threshold of the corresponding segment of the baseline at any position, it is regarded as a potential risk source, and if the curve remains within the threshold, it indicates that the detection environment is stable. The alarm logic is triggered in real time based on this spectrum. The generated complete detection report not only records the excess amplitude, but also uses the spatial position information of the spiral vector index to infer the relative proportion of particle contribution and gas contribution in the abnormal signal, providing operators with targeted tracing ideas. Through this full-link collaboration from the material level to the signal level, this method realizes rapid, synchronous, and online monitoring of submicron particles and volatile gases in laboratory scenarios, significantly reducing the risk of compound pollution caused by particle-gas interaction, and providing real-time, quantitative, and highly sensitive data support for environmental safety management.

[0037] Furthermore, the core principle of time averaging and noise suppression is to treat a raw, pure background signal as a digital sequence containing multi-scale information. Using an equally weighted sliding window, high-frequency perturbations are separated from low-frequency system deformations. Isolated outliers are further removed, leaving the remaining data reflecting only the background fluctuations of the laboratory's static environment. The equally weighted sliding window is essentially equivalent to a linear time-domain low-pass filter: when the window width is set to a one-second window, the algorithm replaces the earliest sample removed with the most recent sample entering the window each time it moves forward by one sampling interval, then calculates a simple arithmetic average of all samples within the window. Because each sample contributes equal weight within the window, this processing does not bias the background amplitude at any particular moment. It also significantly attenuates jitter with periods shorter than one second, preserving true low-speed variations associated with laboratory air exchange, equipment heat dissipation, or minor airflow disturbances. Furthermore, random spikes often appear in the sequence as single points or very short fragments, lacking trend coherence with adjacent samples. The algorithm performs statistical detection on instantaneous amplitude deviations within each second time window: if the amplitude of a sampling point differs from the window average so much that it cannot be caused by slow changes in the environment, it is judged to be a random spike.

[0038] At this point, linear interpolation is used to construct a straight line using the valid sampling values ​​before and after the sampling point and taking its value at the peak position, thereby maintaining the continuity and monotonicity of the signal waveform but removing non-physical jumps. As the window is cyclically shifted, all data undergo the same smoothing and interpolation steps, ultimately resulting in a smooth pure background signal. It has a gentle trend on the time axis and no sharp inflection points on the waveform edge, representing the true noise floor under normal system conditions. Next, the maximum and minimum values ​​of the smooth pure background signal over the entire time range are extracted. The maximum value is defined as the upper threshold, and the minimum value is defined as the lower threshold. The reason for choosing extreme values ​​rather than local statistics is to ensure that the subsequent judgment criteria cover all possible background amplitude ranges: if the real-time signal exceeds the upper threshold, then the amplitude must be higher than any historical background peak; if it is lower than the lower threshold, then it must be lower than any historical background valley. This constructed laboratory environment's particle-gas synergistic signal baseline effectively establishes a safety belt for the system, completely determined by its own historical behavior. This not only avoids bias introduced by external empirical thresholds, but also ensures that if the environmental background drifts continuously during long-term operation, new time averaging and noise suppression processing will recalculate and update the upper and lower thresholds, enabling adaptive tracking. Leveraging this baseline, subsequent multi-segment spiral vector compromise processing only needs to determine whether the real-time signal has breached the safety belt to trigger in-depth analysis and alarm logic. This, from a signal processing perspective, establishes the high sensitivity and low false alarm rate of synergistic detection.

[0039] Furthermore, the blank integrated adsorption carrier plays a key role in establishing a "zero-action" channel for pure background signals, and therefore must be cut and formed as a whole from an inert porous substrate. This integral cutting means that there are no splicing, bonding, or local sintering steps during the preparation process, resulting in naturally extended pore walls and no interfacial fractures, thus avoiding secondary turbulence or particle accumulation induced by micro-gaps. The inert porous substrate does not contain any surface functional groups, which means that the density of its pore wall charge, neutralization sites, and chemical affinity sites approaches zero. As particles and gases pass through it, they are only affected by physical drag and inertia, without chemical adsorption or surface catalytic reactions, thereby minimizing compositional shifts and memory effects during the background sampling stage.

[0040] The continuous, unrestricted pore structure ensures isotropic airflow upon entry into the carrier. The pores lack a specific principal axis and are free of bottlenecks, bends, or blind cavities. This ensures that each microchannel contributes equally to the flow resistance, ultimately creating a stable, laminar flow base with low pressure drop on a macroscale. This provides reproducible fluid dynamic boundary conditions for subsequent replacement of functionalized integrated adsorption carriers within the same flow path. An inert support is employed during installation, strictly controlling flow direction: only the sample to be tested is allowed to enter through a calibrated inlet surface and exit through a single, opposite outlet surface. The inert support mechanically locks the blank integrated adsorption carrier in place, preventing gap leakage caused by thermal expansion or micro-vibration due to external forces. The calibrated inlet surface defines the reference zero point for all subsequent carrier insertions and removals, ensuring consistent flow direction across membrane replacements and experimental batches. The single outlet surface not only prevents external air from diluting the sample at the outlet but also allows for real-time monitoring of the flow field and verification of data integrity during the blank phase, when required, by connecting a mass flow controller or pressure sensor.

[0041] Furthermore, in step 2, a layered amphoteric-philic adsorption film is deposited in situ on the surface of the integrated adsorption carrier to form a micron-scale spiral channel in the adsorption film, comprising: preparing an amphoteric precursor solution and a phobic precursor solution by a proportioning solution method, wherein both precursor solutions are fully homogenized under constant temperature magnetic stirring without introducing any surfactant; then the amphoteric precursor solution and the phobic precursor solution are sequentially packaged in a light-proof container as the only precursor source for in situ deposition; performing in situ deposition, comprising: sequentially depositing an amphoteric precursor solution and a phobic precursor solution on an inert porous substrate; accumulation and staged air drying cycles; each completion of an affinity-phobic bilayer deposition is counted as a cycle, and the total number of cycles is not less than 5 times; the entire in situ deposition process is carried out on a constant rotating platform, and the rotating platform maintains a constant angular velocity of 180 degrees per second at a uniform speed to ensure that the layered affinity-phobic amphoteric adsorption film is evenly covered on the entire surface and has consistent thickness; after the in situ deposition is completed, the rotating magnetic field induction is triggered, and the rotating magnetic field runs in a clockwise direction for a duration of not less than 30 seconds; during the rotating magnetic field induction process, a directional rearrangement occurs inside the layered affinity-phobic amphoteric adsorption film to generate micron-scale spiral channels that run through the thickness of the adsorption film.

[0042] True in-situ deposition occurs on an inert porous substrate. Each cycle of deposition of an affinic precursor solution, aphobic precursor solution, and subsequent air drying forms a pair of layered structures with contrasting chemical properties. The affinic layer preferentially retains polar functional groups during drying and forms a rich hydrogen-bonding network, while the subsequent aphobic layer, due to its molecular chains repelling polar groups, inherently has low surface energy properties. The superposition of the two creates a periodic surface energy gradient through the thickness. The total number of cycles is no less than five, ensuring that this gradient is large enough to cover the diameter of the pores to be formed, thereby ensuring that the subsequent helical rearrangement can simultaneously cross multiple affinic-phobic interfaces with each rotation.

[0043] The entire in-situ deposition process takes place on a rotating platform with a constant angular velocity of 180 degrees per second. Its uniform rotation provides a radial shear field and ensures rapid spreading of the solution on the substrate without forming thickness ripples. As solvent evaporation and air drying alternate, the small molecules within the hydrophilic layer tend to pack more densely due to polar interactions. Driven by its low surface energy, the hydrophobic layer gradually expels residual solvent and contracts slightly. This contraction tension, combined with the hygroscopic expansion of the underlying hydrophilic layer, generates a gradient stress, preserving torsional elastic potential energy for the subsequent shaping of the spiral channel.

[0044] When all cycles are completed and the thickness of the layered hydrophilic-hydrophobic adsorption film reaches the design requirements, a clockwise rotating magnetic field is introduced for at least 30 seconds. The rotating magnetic field provides a macroscopically controllable vector field within the membrane system, causing the magnetic moment to rapidly orient along the plane, thereby inducing the polar molecular chains enriched in the hydrophilic layer to produce synchronous rotation within the plane. The hydrophobic layer, due to its low polarity, hardly responds to the magnetic field and maintains its initial orientation. As the magnetic field continues to drive, the interlayer interface is twisted into a spiral shape and propagates from bottom to top along the thickness direction, eventually generating continuous micron-scale spiral channels throughout the thickness of the membrane. The surface energy difference brought about by the alternation of hydrophilic-hydrophobic layers determines that the pore wall presents a periodic distribution of hydrophilic and hydrophobic regions at the same time, resulting in an energy landscape of alternating hydrophilic and hydrophobic bipolarity within the channel. This can temporarily adsorb polar gas molecules with high affinity, while reducing the desorption resistance of hydrophobic gases with low surface energy. The spiral geometry also applies lateral inertial displacement to submicron particles, realizing the physical field support for the simultaneous enrichment of particles and gas. At the same time, the directional rearrangement induced by the rotating magnetic field also imparts a single-handed geometric chirality to the pores. This chirality, under the action of the pulsed airflow, generates a combined effect of Coriolis, centrifugal, and shear forces, forming a spiral flow field with axial propulsion and radial stratification, providing a hydrodynamic channel for the rapid extraction of substances from the adsorbed layer during the subsequent optical-acoustic dual-path excitation stage. Thus, the layered amphoteric adsorption membrane's three functionalities—chemical heterogeneity, geometric chirality, and fluid response—are simultaneously activated. This completes the functional upgrade of the substrate surface from a materials engineering perspective and lays a decisive foundation for the subsequent simultaneous enrichment, collaborative analysis, and anomaly determination of laboratory particle-gas collaborative detection methods.

[0045] Furthermore, the internal torsional elastic properties of the micron-scale spiral channels generated during the gyromagnetic traction process are fully relaxed, and the microscopic wrinkles formed by the interfacial tension difference between the affinity layer and the hydrophobic layer are also rearranged under the control of the surface tension, so that the spiral curvature of the channel wall tends to static equilibrium, thus providing a stable geometric foundation for subsequent heat treatment. The membrane layer and the inert porous substrate are then sent into a vacuum heat curing chamber. The temperature is gradually raised to 120 degrees Celsius under negative pressure and maintained for 45 minutes. This process not only uses the low-pressure environment to lower the boiling point of the solvent and trace water, allowing the residual volatile components to completely escape, avoiding the formation of endogenous bubbles that damage the channels during the heating process, but also provides thermal energy drive for the reactive groups on the affinity-hydrophobic interface, allowing them to undergo cross-linking and curing under the molecular orientation pairing induced by the polarity difference. Cross-linking and curing transforms the interlayer structure originally maintained by van der Waals forces and hydrogen bonds into a covalent network, significantly improving the mechanical stiffness and thermal stability of the layered amphoteric-philic adsorption membrane. At the same time, a gradient cross-linked shell is formed on the pore surface, which not only locks the chemical heterogeneity of the amphoteric-philic interface, but also prevents interlayer delamination caused by subsequent pulse airflow impact.

[0046] During the cooling process to room temperature, the chamber is maintained under negative pressure, preventing external water vapor from infiltrating the pores during the cooling phase. Otherwise, polar water molecules would preferentially occupy the affinitive sites, weakening the adsorption selectivity for the target gas molecules and potentially triggering stress concentration and microcracks during repeated temperature cycles. After thermal curing, the cross-linked, layered, amphoteric adsorption membrane is integrally encapsulated with a prefabricated multi-stage impactor collection structure using an isobaric interlocking method. This isobaric approach applies uniform normal pressure across the interface between the membrane and the impactor collection structure, ensuring a simultaneous, close fit across the entire interface without excessive local compression, thus preventing pore deformation or membrane warping caused by pressure differentials. The inlet surface of the multi-stage impactor collection structure is kept strictly parallel to the outer surface of the adsorption membrane, enabling the particle size classification and spiral pores to work synergistically. During actual testing, the impactor stage captures larger particles, followed by the pores for simultaneous enrichment of submicron particles and volatile gases. A circumferential inert sealing gasket fills any microgaps under isobaric loading, blocking bypass airflow and ensuring that all tested samples pass through the designed path of the functionalized integrated adsorption carrier. Thus, a functionalized integrated adsorption carrier with chemical stability, mechanical integrity, and fluid uniformity has been formed, providing an efficient enrichment platform and reliable signal generation source for subsequent particle-gas collaborative detection.

[0047] Furthermore, during the packaging process, the inlet surface of the multi-stage impactor collection structure is maintained strictly parallel to the outer surface of the layered amphoteric-philic adsorption membrane to ensure a uniform velocity distribution across the cross section of the sample flow entering the functionalized integrated adsorption carrier. Any angular deviation will cause a localized change in the angle of attack, which in turn will form a short-range shock wave and turbulent zone on the relatively high-pressure side. This will cause the particle inertial excursion trajectory to misalign with the designed capture path of the spiral channel, ultimately reducing the simultaneous enrichment efficiency. In engineering, a precision alignment jig is used to fix the reference surfaces of the two structures in the same reference plane. Laser flatness measurement is then used to confirm that the height difference across the entire contact area is less than one percent of the channel diameter, thereby eliminating tilt at the macroscale. Micron-thick compensation gaskets are then used to adjust the remaining error to ensure that there is no air gap between the two. If microgaps exist, the airflow will choose a bypass path with less resistance, causing some of the sample to bypass the spiral channel and multi-stage impactor collection structure and exit directly without enrichment. This dilutes the real-time signal amplitude and introduces difficult-to-filter random fluctuations in the frequency domain. In order to completely block potential leakage paths, inert sealing gaskets are added circumferentially to the assembly interface. The inert sealing gaskets produce radial rebound after axial compression, achieving continuous circumferential sealing of the entire circumference to avoid airflow bypass.

[0048] The inert sealing gasket is made of chemically inert materials, such as high-temperature sintered polytetrafluoroethylene or fluorosilicone rubber. This ensures that volatile gas components are neither adsorbed nor released, while exhibiting low compression set, maintaining sealing force even after repeated thermal cycling and pulsed airflow. During installation, an isobaric loading ring is used. Uniform normal pressure is applied via symmetrical bolts or pneumatic clamps with preset torques, ensuring that the inert sealing gasket reaches the designed compression ratio simultaneously along its circumference, preventing localized overpressure from causing channel deformation or shear damage. After packaging, seal integrity can be verified using differential pressure leak testing or helium mass spectrometry. When no leaks are confirmed and the flow resistance curve matches the blank carrier calibration value, the multi-stage impact collection structure and the layered amphoteric adsorption membrane form a gap-free, circumferentially sealed structure. The resulting functionalized integrated adsorption carrier ensures that all sample gas flows through the spiral channel according to the specified flow path and velocity field, without signal attenuation or error bias due to structural assembly defects. This provides reliable fluid dynamic boundaries and airtightness for subsequent particle-gas collaborative detection.

[0049] Furthermore, in laboratory particle-gas synergy detection methods based on aerosol adsorption mechanisms, multi-segment spiral vector compromise processing is responsible for converting the raw particle-gas synergy signal into a highly discriminative particle-gas synergy signature. First, the system divides the complete time series data into segments of equal length according to the triggering order of the photoacoustic pulses. Each segment maintains closed boundaries on the time, amplitude, and phase axes, thereby ensuring that there is no cross-segment leakage during subsequent comparisons. The algorithm then calculates a spiral vector index for each segment. This index maps multidimensional features within the segment, such as peak amplitude, rising edge steepness, half-width, and phase delay, into the same vector space, geometrically ensuring that the particle and gas contributions share a set of coordinates. To prevent reverse crossover between adjacent segments during priority evaluation, the system performs a crossover-free interpolation ranking of all spiral vector indices: using the segment number as the initial reference, the indices are then interpolated based on the overall similarity between the vectors until a monotonically increasing segment priority list is obtained.

[0050] Next, the dynamic threshold adjustment phase begins. The algorithm uses the segment priority list as the order, starting from the starting point of the safety belt corresponding to the baseline, and performs progressive threshold updates segment by segment. Each update subtracts the current segment spiral vector index from the previously confirmed safe threshold state. When the difference exceeds the edge of the safety belt, the new threshold is moved up or down, and the deviation at this step is written into the heterogeneous reconstruction matrix. After a complete traversal, the heterogeneous reconstruction matrix records the entire offset trajectory of each segment relative to the baseline, but it still contains isolated high-frequency noise entries. To this end, the system calls the compromise fusion rule, scanning forward and backward in sequence while maintaining the priority order unchanged. Matrix entries that lack trend consistency with adjacent segments and whose deviation amplitude is below the statistical significance threshold are bidirectionally pruned, and the final heterogeneous reconstruction matrix is ​​obtained after elimination. The final step maps the final heterogeneous reconstructed matrix into a particle-gas synergistic characteristic map: the horizontal axis presents a time series by segment number, while the vertical axis displays a multidimensional characteristic line projected by spiral vector index. Each significant boundary violation in the map corresponds to an abnormal amplification of particles or gas components during the synchronized release process. The system uses this information to quickly locate the risky segment and trigger the alarm logic. Through this end-to-end data processing chain, the complex fluctuations of the original signal are compressed into visual, comparable, and traceable spectral information. This not only preserves the time-frequency-amplitude characteristics of particle-gas synergistic release, but also minimizes false triggers caused by environmental noise and instrument drift, providing a stable and reliable real-time judgment for laboratory environmental safety monitoring.

[0051] Furthermore, during the data parsing phase of a laboratory particle-gas collaborative detection method based on aerosol adsorption, a key step in the multi-segment spiral vector compromise process is to construct a complete, monotonically increasing segment priority list. The core logic for generating this list involves sequentially inserting the spiral vector index of each segment into the existing sequence through interpolation. In implementation, the system first selects an initial baseline segment from the entire segment set, either randomly or by a preset rule, and directly places the spiral vector index of this initial baseline segment at the beginning of the segment priority list, marking it as the baseline index. While the selection of the baseline index may seem arbitrary, since the spiral vector index itself carries multidimensional features such as amplitude, phase, and half-width, it naturally resides at a measurable position in the entire vector space. Therefore, using any segment as a starting point ensures that subsequent interpolation position calculations maintain a consistent reference coordinate within the metric space. Next, the algorithm performs the interpolation position determination operation for each unranked segment: first, the vector space Euclidean distance between the spiral vector index of the current segment to be ranked and the spiral vector indices of all ranked segments in the segment priority list is calculated.

[0052] The significance of Euclidean distance is to measure the difference in multi-dimensional features with a unified dimension. The smaller the distance, the closer the overall performance of the two segments in the coordinated release behavior, and the larger the distance, the more obvious the difference in the kinetic patterns of the two segments during synchronous enrichment or release. After calculating all the distances, the algorithm sorts these distances from small to large, and then determines that the spiral vector index of the segment to be sorted should be inserted between the two closest sorted segments. The smallest distance provides the upper bound, and the second smallest distance provides the lower bound. At this time, the system automatically determines the relative position of the segment to be sorted relative to the above two reference points based on the size of the Euclidean distance, thereby obtaining a clear insertion rank position. In order to ensure that the segment priority list is always monotonically increasing during the insertion process and there is no cross-contradiction, the algorithm will verify the order consistency of the old and new indexes in the vector space before insertion: once it is found that the insertion operation may cause the previously determined segment priority to be reversed or parallel crossed, the system will re-evaluate the selection of the nearest neighbor according to the distance weight until a unique insertion rank position that satisfies both the monotonicity and distance minimization principles is found. With each rank insertion, the segment priority list is expanded by one entry, and the relative order of all sorted segments in the vector space is immediately locked, no longer changing due to subsequent insertions. The algorithm repeats this operation for the remaining segments until all segment spiral vector indices have a clear insertion position relative to the existing segments in the segment priority list. Finally, all segment spiral vector indices are inserted into the corresponding positions in sequence, resulting in a complete, monotonically increasing segment priority list.

[0053] Because each step is based on a real-valued metric of Euclidean distance and exhaustively enumerates minimum-distance rank-interpolation schemes under constraints, the list inherently minimizes the length of the sequence path within the high-dimensional vector space. From an information-theoretic perspective, this reduces the state space that must be traversed for subsequent threshold updates and smooths threshold drift during progressive threshold updates. More importantly, this monotonically increasing segment priority list physically corresponds to the trend sequence of particle-gas synchronous release events: segments closer to the front represent release behaviors most similar to the baseline index, while segments closer to the back deviate more significantly from the baseline index. When the algorithm further utilizes this list to compare against the baseline, perform progressive threshold updates, and generate a heterogeneous reconstruction matrix, segments with higher priority have a greater influence on threshold changes due to their high similarity to the baseline index. Lower priority segments, however, only significantly influence matrix entries after threshold adaptation, thereby achieving adaptive hierarchical screening of anomalous signals. The final mapped particle-gas synergistic characteristic map inherits the monotonicity of the segment priority list in both the time dimension and the characteristic dimension, so that any abnormal peak is not only prominent in amplitude, but also shows a clear distance from the benchmark index in the path sequence, thereby providing clear and traceable risk location clues and providing high-precision and high-reliability judgment criteria for laboratory environmental safety monitoring.

[0054] Furthermore, after all segment spiral vector indices are written into the segment priority list according to the insertion rank rule, whenever two consecutive indices are extracted and their Euclidean distances are calculated, the distance between the latter and the previous one must be greater than the distance between the previous pair of adjacent indices. This cycle repeats until the end of the list, resulting in the finding that the Euclidean distance between the spiral vector indices of any two adjacent segments shows a gradually increasing trend. This physical implication can be understood as the coordinated release behavior following a continuously expanding, divergent trajectory within a high-dimensional feature space: segments closer to the starting point of the list are closer to the baseline index morphology, while segments closer to the end of the list exhibit more significant multidimensional deviations from the baseline index. Because the spiral vector index integrates multiple dimensions such as peak amplitude, rising edge steepness, half-maximum width, and phase delay, the absolute magnitude of the Euclidean distance directly represents the global differences between the segment release dynamics patterns. As these differences increase between adjacent segments, the priority path is geometrically free of loops or turns, thus preventing any "cross-contradictions." This zero-return trajectory has three implications: first, progressive threshold gate updates can rely on a distance sequence that neither converges abruptly nor oscillates locally to smoothly adjust the threshold, thus avoiding multiple local flips at the signal boundary that could lead to misjudgment; second, the heterogeneous reconstruction matrix can use a fixed incremental window when performing bidirectional pruning based on priority, eliminating the need to recalculate the threshold for local jumps, thereby significantly reducing the amount of real-time computation; finally, when the particle-gas collaborative characteristic map maps this divergent path onto a two-dimensional graph in the visualization stage, the curve maintains a unidirectional discrete expansion, and any over-threshold peak will give a clear amplitude difference in the vertical space and an irreversible process direction on the horizontal time axis, making the abnormal section not only easy to locate, but also its severity can be intuitively quantified by the distance from the benchmark index in the vector space. Strict adherence to the monotonically increasing definition of "gradually increasing Euclidean distance" essentially creates a barrier-free logical channel for the entire collaborative detection process at the data structure level. This ensures that each step, from signal generation, segmentation, index extraction, priority sorting, threshold gate updating, matrix cropping, and map rendering, is based on the stable and irreversible geometric order relationship of the previous step. This allows the system to maintain the continuity and discriminative power of the particle-gas simultaneous enrichment signal in the face of multi-source noise, occasional drift, or extreme operating conditions, providing real-time judgment criteria with sufficient traceability and statistical robustness for laboratory environmental safety monitoring.

[0055] Furthermore, the section priority list refers to the list of sections numbered S1, S2, ..., S obtained by sorting without cross-interpolation. N And all the segment sequences that satisfy the monotonically increasing relationship. Particle-gas synergistic signal baseline: The upper and lower threshold sets defined by the smooth pure background signal obtained in step 1, denoted as Where k is the feature dimension number and K is the total number of feature dimensions. Threshold vector: The dynamic upper and lower bounds that act on all feature dimensions simultaneously in each progressive cycle, denoted as The subscript i corresponds to the segment S i Heterogeneous reconstruction matrix: An N×2K matrix formed by concatenating the threshold gate vectors output by all progressive loops, denoted as R.

[0056] Progressive threshold update, with the segment priority list {S1,…,S N} in sequence, and compared with the particle-gas cooperative signal baseline B, specifically including: setting the progressive loop starting index i←0. Using the baseline to define the initial threshold vector

[0057] Perform the following sub-steps for i=1 to N in sequence until all segments are processed. i Extract the same-dimensional feature vector from the corresponding original particle-gas cooperative signal Calculate the deviation vector between the segment feature vector and the baseline center value Segment S according to the priority order i Assign weight coefficient where λ i ∈(0,1] increases monotonically with increasing priority and is used to control the convergence rate of the threshold.

[0058] Using the deviation vector and the progressive weight coefficient, the threshold vector Θ of the previous cycle is i-1 Perform the update: in: For segment S i The variance estimate is obtained by the fast discrete variance estimation algorithm on the feature dimension k; β∈[0,1] is a fixed safety buffer coefficient used to limit the threshold convergence amplitude; the above formula ensures that when the deviation vector is positive / negative, the threshold vector is progressively adjusted outward / inward respectively to form an adaptive segment threshold.

[0059] To prevent the threshold vector from exceeding the physical limit, a robustness correction is performed on the update result:

[0060] Where η>0 is the width of the baseline extension redundant domain, which ensures that the updated threshold gate always drifts around the baseline in a limited manner.

[0061] Write the corrected threshold vector into the heterogeneous reconstruction matrix row by row: When i=N, all segments have completed the progressive threshold gate update. It is fully written, and each row of the matrix corresponds to the final threshold vector of the section with the same serial number in the priority list. The completeness of the matrix is verified by the following condition: for any i < j and any k, it always holds that This ensures that the threshold expands progressively as the priority increases. The number of rows in the matrix is equal to the total number of sections N, and the number of columns is equal to 2K, without missing or duplicate rows and columns.

[0062] Furthermore, a priority weight w is assigned to each row i i ; the weight monotonically increases as the section priority increases. The weight function does not depend on the threshold value itself and is only determined by the section serial number, ensuring the consistency of the overall sorting. For any row i and feature dimension k in R, the threshold span is calculated The row stability D i is defined as the arithmetic mean of all K threshold spans. A small span indicates that the threshold converges stably. A compromise score P is defined for each row i , which is given by the linear superposition of the "low-priority penalty term" and the "high-span penalty term"; the higher the compromise score, the smaller the contribution of the row to the overall discrimination and the more inclined it is to be deleted. The weight coefficients of the two penalty terms are set in the initial calibration stage of the system to ensure the repeatability of the deletion strategy. The column variance V is calculated for each column j ; too small variance indicates that the information gain of the column in all retained rows is limited.

[0063] According to the priority weight function, a row weight vector {w1,..., w N [[ID=2)]]} is generated. The D of each row is calculated according to the threshold stability index i and the compromise score P is obtained i . The compromise scores are sorted in descending order to form a sequence of candidate rows to be deleted. Set the maximum number of rows that can be deleted R max (a calibration parameter, which can be given by the baseline redundancy). Starting from the head of the sequence of candidate rows to be deleted, check the rows S in turn m : If P m exceeds the compromise score threshold and the current number of deleted rows has not reached R max , then delete this row; if either condition is not satisfied, stop the outward pruning. After completion, a temporary matrix R' with N' rows is obtained, and the retained rows still maintain the original priority order.

[0064] Calculate the variance V for all columns of R' j to obtain the column significance vector. Set the minimum significance threshold V min (a calibration parameter for ensuring that the remaining columns have discrimination for risk discrimination). Sort the column significance vector from low to high to form a sequence of candidate columns to be deleted. Check the columns C in order n : If then delete this column and its paired column (the lower threshold column and the upper threshold column are a pair) as a whole; if Then stop the inward pruning. After completion, we get a matrix R″ with 2K′ columns. The column order still maintains the original arrangement rule of “lower threshold gate first, then upper threshold gate”. Make sure that the number of rows in R″ is not less than the minimum number of rows N. min If it is insufficient, roll back the most recent row deletion. Ensure that each feature dimension of R″ still retains paired threshold columns, and the total number of feature dimensions K′≥K min If not enough, roll back the most recent column deletion.

[0065] The following is a complete set of specific embodiments for implementing a laboratory particle-gas collaborative detection method based on an aerosol adsorption mechanism. The values ​​listed are examples and can be appropriately scaled according to actual needs. Substrate material: polytetrafluoroethylene; cylindrical diameter d0 = 50 mm, thickness h0 = 10 mm, porosity φ0 = 0.70, average pore size r0 = 20 μm. Affinity precursor solution concentration C p =0.05mol·L -1 ; Concentration of hydrophobic precursor solution C h =0.03 mol·L -1 ; Single hydrophobic double layer deposition thickness t cycle =2μm cycle number n cycle =5, final film thickness t film =10μm; angular velocity of the rotating platform ω = 180°·s -1 =πrad·s -1 Magnetic induction intensity B ind =0.05T; duration t B =30; after induction, the pitch of the micron-scale spiral channel p = 5 μm, and the channel diameter d pore =1μm.

[0066] Cavity pressure P vac =500Pa; heating rate v T =2K·min -1 ; to T max =393K; holding time t cure =45min. Carrier gas: nitrogen; volume flow rate Q = 2L·min -1 ; Particles: sodium chloride aerosol, count concentration C part =5×10 4 cm -3 , median diameter d med =300nm; Volatile component: ethanol vapor volume fraction C eth =50ppm. Duty cycle 50%, frequency f pulse =2Hz; valve opening time t on =250ms, valve closing time t off =250ms; total number of pulses N pulse =120.

[0067] Acquisition duration T bg =60s, discrete sampling rate f s =1000Hz. The original background voltage sequence is recorded as {u0(1),u0(2),…,u0(6×10 4 )};Equal weight sliding window width T w =1s. Number of samples in the window N w =T w ·f s = 1000. The average value of the mth window is

[0068] The random spike determination threshold is where σ m is the standard deviation within the window. The peaks that fall outside the threshold are replaced by linear interpolation to obtain a smooth background sequence

[0069] The baseline upper and lower threshold vectors B=(B min ,B max ). On the surface of the inert porous substrate, a cycle of affinity solution → air drying → hydrophobic solution → air drying is considered. After 5 cycles, the total film thickness reaches 10 μm. Ensure that the solution is spread evenly. After 30 seconds of rotating magnetic field induction, spiral channels are formed; the channel density

[0070] Single pulse volume Average residence time of gas in the pores Approximate enrichment fold The enrichment factor after ten or twenty pulses is accumulated as E cum =E 120 ≈1.44 120 ≈3.6×10 25 The sensitivity far exceeds the detector's requirements (subsequent threshold gates will converge dynamically). Optical-acoustic dual-path excitation, synchronous output of three-dimensional features: amplitude A, spectral centroid F c , rise time T r . Calculate x for each pulse segment i =(A i ,F c,i ,T r,i ), i = 1, ..., 120. Combine 120 pulses into 3 segments, with segment length L = 3, and a total of N = 40 segments. Define the spiral vector index for segment i as Where μ represents the mean value within the segment. After mapping to polar coordinates, the priority list S1 of the segments is obtained by sorting according to the angle of argument and interpolating without cross-ranking. <S2<…<S 40 .

[0071] Baseline Center For the jth segment, let β=0.50. For feature k∈{1,2,3}(corresponding to A,F c ,T r )implement

[0072] in Initial value Get the threshold gate vector rows from j = 1 to 40 to form a matrix P j =α(1-w j )+γD j , Take α = 0.60, γ = 0.40. Delete the top 10 rows and keep N′ = 30. Column variance threshold V min =10 -6 All column variances are higher than the threshold, so the number of columns remains 6. The final heterogeneous reconstruction matrix is ​​obtained Linearly map R″ to the pseudo-color space to obtain the particle-gas synergistic characteristic spectrum In the amplitude dimension of segment i=7, The system determines that the "feature exceeds the threshold", triggers the alarm logic, and generates a data including the timestamp, the exceeded dimension, and the exceeded amplitude ratio: A complete test report will be provided; and the mass spectrum of volatile components will be automatically recorded during the same period for subsequent tracking. -1 Under the flow rate, the system can achieve 10 3 The fusion matrix is ​​enriched to push the detection limit of 50ppm ethanol down to 0.5ppm. The overall detection delay is ≤15s, meeting the rapid early warning needs of laboratories. The compromised fusion achieves a 25% data compression rate, significantly reducing the computational and storage burden compared to the original matrix while maintaining >98% feature fidelity.

[0073] Furthermore, after the functionalized integrated adsorption carrier completes the synchronous enrichment, submicron particles and various types of volatile gas molecules are retained in the micron-scale spiral channels in the adsorption membrane at the same time. At this time, the system starts the optical-acoustic dual-path excitation module. First, the optical path applies a directional pulsed laser or a broadband heat source to the surface of the adsorption membrane. The optical excitation causes the local micro-area to heat up in a very short time. Especially in the affinity layer, the polar groups are more sensitive to thermal radiation and are quickly transmitted to the inside of the channel. This thermal disturbance causes the desorption rate of gas molecules in the membrane layer to increase significantly, and the volatile components that were originally bound to the interface in the form of physical adsorption or weak hydrogen bonds are rapidly desorbed, forming a primary gas release peak. At the same time, the local temperature rise will also cause thermal expansion and contraction effects in the membrane layer, driving the particles in the channel to produce slight displacements, providing a prerequisite for the particles to participate in signal formation.

[0074] Next, the acoustic path is excited and activated, usually by a piezoelectric transducer emitting an acoustic wave of a specific frequency, forming a periodic acoustic pressure gradient. When the acoustic wave propagates in the spiral channel, it produces a coupling effect with the spiral structure of the channel, causing the acoustic pressure to superimpose in the axial and radial directions to form a miniature standing wave field. Under the disturbance of the acoustic field, the gas molecules are accelerated and driven away from the residual adsorption sites on the pore wall, enhancing the desorption integrity; the particles, because their mass and size are significantly larger than those of gas molecules, experience considerable acoustic-driven motion in the acoustic pressure gradient, manifesting as slight vibrations or secondary release along the channel direction. This particle response mechanism not only increases the probability of particle release to the outside world, but also gives its release process a strong time-marking characteristic under the modulation of the acoustic wave frequency.

[0075] Because optical excitation primarily affects the thermal desorption behavior of the gas, while acoustic excitation affects both gas disturbance and particle micromotion, the two mechanisms work synergistically within the same time period, causing particle release and gas release to overlap in time. At the same time, their respective release intensities complement each other, ultimately forming a composite original particle-gas synergistic signal at the sensor output. This signal exhibits typical multi-peak pulse characteristics, where the amplitude, morphology, width, and response time between different peaks reflect the enrichment, binding strength, and release path of the particles and gas components in the pores. By further performing segmentation and vector analysis on the original signal, multi-dimensional synergistic information with structured characteristics can be obtained to support subsequent anomaly identification and map mapping processes.

[0076] Figure 2Schematic diagram of the three-dimensional structure of micron-scale spiral channels within the integrated adsorption carrier. As shown in the figure, the functionalized integrated adsorption carrier has an overall rectangular structure, consisting of a bottom layer of an inert porous substrate and an upper layer of a layered amphoteric-philic adsorption membrane. The inert porous substrate is cut and formed from an inert porous material. The substrate does not contain any surface functional groups, and its continuous pore structure remains unrestricted and provides stable structural support for the upper adsorption membrane. The layered amphoteric-philic adsorption membrane is formed via an in situ deposition process and comprises alternating amphoteric and amphoteric layers. The amphoteric layer is deposited from an amphoteric precursor solution, and the amphoteric layer is deposited from amphoteric precursor solution. Both precursor solutions are thoroughly homogenized under constant temperature magnetic stirring. The deposition process is performed on a constantly rotating platform, which rotates at a constant angular velocity of 180 degrees per second to ensure uniform coverage and thickness of the layered structure across the entire surface. Each completed amphoteric-philic bilayer deposition is counted as one cycle, with a total of at least five cycles. After being induced by a rotating magnetic field, the layered amphoteric-philic adsorption membrane undergoes a directional rearrangement, generating micron-scale spiral channels that penetrate the thickness of the adsorption membrane. These spiral channels are distributed in a three-dimensional spiral shape, with a channel diameter ranging from 1 to 5 microns, providing an effective enrichment channel for submicron particles and volatile gas components. The airflow enters from the calibrated air inlet surface on the left side of the carrier, passes through the separation and enrichment effect of the spiral channels, and is discharged from the opposite single air outlet surface on the right side. The design of the spiral channel structure enables particles of different sizes and gas components of different properties to be effectively separated and simultaneously enriched during the passage process, providing ideal sample pretreatment conditions for subsequent photoacoustic dual-path excitation detection.

[0077] Figure 3 This is a schematic diagram of the simultaneous enrichment process of submicron particles and volatile gas components. The figure uses a time-concentration coordinate system, with the horizontal axis representing the enrichment time in seconds, ranging from 0 seconds to 7 seconds; the vertical axis represents the enrichment concentration in μg / m 3 , concentration range from 0 to 140 μg / m 3The figure contains two enrichment curves: the solid line represents the submicron particle enrichment curve, and the dotted line represents the volatile gas component enrichment curve. Under pulse mode operating conditions, the sample gas flow to be tested passes through the functionalized integrated adsorption carrier at a set pulse frequency. The pulse mode parameters are set as follows: pulse frequency 0.5 Hz, pulse duration 1 second, and interval time 1 second. This pulse mode ensures that the sample has sufficient residence time in the spiral channel for effective enrichment. It can be observed from the enrichment curves that submicron particles and volatile gas components exhibit different dynamic characteristics during the enrichment process. The submicron particle enrichment curve shows that the concentration rises rapidly in the early stage of enrichment, reaching a high concentration level at 3-4 seconds, and then the upward trend gradually slows down. The volatile gas component enrichment curve shows a similar trend, but the enrichment rate is slightly lower than that of submicron particles. This difference is mainly due to the different transport mechanisms of particles and gas components in the spiral channel: particles are mainly enriched through physical interception and inertial separation, while gas components are mainly enriched through the selective adsorption of the adsorption membrane. The synchronous upward trend of the two curves proves that the micron-scale spiral channels of the layered amphoteric-philic adsorption membrane have good synchronous enrichment capabilities for submicron particles and volatile gas components, laying the foundation for the realization of particle-gas collaborative detection.

[0078] Figure 4This figure compares the particle-gas synergistic characteristic spectrum with the baseline threshold. This figure establishes a complete signal intensity detection coordinate system. The horizontal axis represents the detection time segment, ranging from 1 to 12 segments; the vertical axis represents signal intensity in relative units, ranging from 0 to 70 relative units. Three key reference lines are set in the figure: the baseline is located at a signal intensity of 20 relative units, the upper threshold is located at 50 relative units, and the lower threshold is located at 30 relative units. These thresholds are determined by time averaging and noise suppression of the pure background signal. The upper threshold corresponds to the maximum value of the smoothed pure background signal over the entire time range, and the lower threshold corresponds to the minimum value over the entire time range. The figure shows particle-gas synergistic characteristic spectra under two different conditions. The dotted line represents the normal state characteristic spectrum. This curve fluctuates slightly around the baseline throughout the detection process, and all data points fall within the upper and lower threshold ranges, indicating a safe detection environment. The solid line represents the abnormal state characteristic spectrum. This curve shows a significant drop in signal intensity in detection segments 3, 4, and 5, exceeding the upper threshold range, triggering the alarm logic. The heterogeneous reconstruction matrix mapping process divides the original particle-gas synergistic signal into equal-length segments according to the pulse sequence, and generates a spiral vector index for each segment. Through multi-segment spiral vector compromise processing, the spiral vector index of each segment is sorted without cross-ranking to obtain a monotonically increasing segment priority list. Based on the progressive threshold gate update result, the compromise fusion rule is called to perform bidirectional trimming on the heterogeneous reconstruction matrix, and finally a particle-gas synergistic feature map is generated. When the detection system finds that any feature exceeds the baseline threshold, it automatically triggers the alarm logic and generates a complete detection report, which indicates the location of the over-threshold segment and the alarm level. If all features are within the baseline threshold, the safety status report is output and archived to ensure the reliability and continuity of laboratory environmental monitoring.

[0079] The present invention is not limited to the aforementioned specific embodiments, but extends to any new features or any new combination disclosed in this specification, as well as any new method or process steps or any new combination disclosed.

Claims

1. A laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism, characterized by: The method comprises: Step 1: The sample airflow to be tested is transported at a constant speed through a blank integrated adsorption carrier to collect a pure background signal. The baseline of the particle-gas synergistic signal in the laboratory environment is established by time averaging and noise suppression of the pure background signal. Step 2: In situ depositing a layered adsorption film with both affinity and repellency on the surface of the integrated adsorption carrier to form micron-scale spiral channels in the adsorption film; encapsulating the adsorption film with a multi-stage impact collection structure to obtain a functionalized integrated adsorption carrier for laboratory particle-gas collaborative detection; Step 3: The sample gas flow to be tested is passed through the functionalized integrated adsorption carrier in a pulsed mode. The micron-scale spiral pores of the layered amphoteric-philic adsorption membrane are used to achieve simultaneous enrichment of submicron particles and volatile gas components. After the simultaneous enrichment is completed, the original particle-gas synergistic signal is obtained through optical-acoustic dual-path excitation. Step 4: Perform multi-segment spiral vector compromise processing on the acquired original particle-gas synergistic signal, and map the final heterogeneous reconstruction matrix into a particle-gas synergistic feature map; compare the particle-gas synergistic feature map with the baseline one by one; if any feature exceeds the baseline threshold, the alarm logic is automatically triggered and a complete detection report is generated; if all features are within the baseline threshold, a safety status report is output and archived.

2. The laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism according to claim 1, characterized in that: In step 1, the obtained pure background signal is subjected to time averaging and noise suppression processing over the entire time range; the time averaging and noise suppression processing uses an equal-weight sliding window method to fully smooth the pure background signal according to a one-second time window, and linear interpolation is performed on the random spikes in each time window; after the processing is completed, a smoothed pure background signal is obtained; the maximum value of the smoothed pure background signal over the entire time range is used as the upper threshold, and the minimum value of the smoothed pure background signal over the entire time range is used as the lower threshold to construct the particle-gas synergistic signal baseline of the laboratory environment.

3. The laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism according to claim 2, characterized in that: The blank integrated adsorption carrier is cut and formed as a whole from an inert porous substrate. The inert porous substrate does not contain any surface functional groups, and the continuous pore structure remains through and has no directional restrictions. When the blank integrated adsorption carrier is installed on the inert bracket, it only allows the sample airflow to be tested to enter from the calibrated air inlet surface and be discharged from the opposite single air outlet surface.

4. The laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism according to claim 3, characterized in that: In step 2, a layered amphoteric-philic adsorption film is in situ deposited on the surface of the integrated adsorption carrier to form a micron-scale spiral channel in the adsorption film, which includes: preparing an amphoteric precursor solution and a phobic precursor solution by a proportioning solution method, and both precursor solutions are fully homogenized under constant temperature magnetic stirring without introducing any surfactant; then the amphoteric precursor solution and the phobic precursor solution are sequentially packaged in a light-proof container as the only precursor source for in situ deposition; performing in situ deposition, including: sequentially depositing an amphoteric precursor solution, a phobic precursor solution and a staged deposition on an inert porous substrate. Air drying cycle; each completion of the affinity-phobic bilayer deposition is counted as one cycle, and the total number of cycles is not less than 5 times; the entire in-situ deposition process is carried out on a constant rotating platform, and the rotating platform maintains a constant angular velocity of 180 degrees per second at a uniform speed to ensure that the layered affinity-phobic amphoteric adsorption film is evenly covered on the entire surface and has a consistent thickness; after the in-situ deposition is completed, the rotating magnetic field induction is triggered, and the rotating magnetic field runs in a clockwise direction for a duration of not less than 30 seconds; during the rotating magnetic field induction process, a directional rearrangement occurs inside the layered affinity-phobic amphoteric adsorption film, generating micron-scale spiral channels that penetrate the thickness of the adsorption film.

5. The laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism according to claim 4, characterized in that: In step 2, the adsorption membrane and the multi-stage impact collection structure are integrally packaged to obtain a functional integrated adsorption carrier for laboratory particle-gas collaborative detection. The process includes: after the rotating magnetic field is released, it is allowed to stand for three minutes to allow the micron-scale spiral channel morphology to spontaneously stabilize; the layered amphoteric and amphoteric adsorption membrane forming micron-scale spiral channels is placed together with the inert porous substrate into a vacuum thermal curing cavity; under a negative pressure state, the temperature is gradually raised to 120 degrees Celsius and maintained for 45 minutes to allow the amphoteric and amphoteric interface to undergo cross-linking and curing. After the cross-linking and curing is completed, the temperature is lowered to room temperature, and the cavity is maintained at a negative pressure during this period to prevent external water vapor from entering; the cross-linked and cured layered amphoteric and amphoteric adsorption membrane and the prefabricated multi-stage impact collection structure are integrally packaged by isobaric interlocking.

6. The laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism according to claim 5, characterized in that: During packaging, the air inlet surface of the multi-stage impact collection structure is ensured to be parallel to the outer surface of the layered amphoteric-philic adsorption membrane, and there is no air gap between the two; an inert sealing gasket is used to achieve circumferential sealing to avoid airflow bypass.

7. The laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism according to claim 6, characterized in that: In step 4, the obtained original particle-gas synergistic signal is subjected to multi-segment spiral vector compromise processing, and the process of mapping the final heterogeneous reconstruction matrix into a particle-gas synergistic characteristic map includes: dividing the original particle-gas synergistic signal into equal-length segments according to the pulse sequence, generating a spiral vector index for each segment, and uniquely identifying it with the segment number; performing cross-rank sorting on the spiral vector index of each segment to obtain a monotonically increasing segment priority list; comparing the baseline in the order of the segment priority list, performing progressive threshold gate update on each segment to form a heterogeneous reconstruction matrix; based on the progressive threshold gate update result, calling the compromise fusion rule, bidirectionally trimming the heterogeneous reconstruction matrix according to the priority, and obtaining the final heterogeneous reconstruction matrix; mapping the final heterogeneous reconstruction matrix into a particle-gas synergistic characteristic map.

8. The laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism according to claim 7, characterized in that: An initial base segment is randomly selected from all segments, and the spiral vector index of the initial base segment is placed as the base index at the starting position of the segment priority list; for each of the remaining segments, the insertion rank position between its spiral vector index and the spiral vector index of each segment in the current segment priority list is calculated respectively, wherein the insertion rank position determination process includes: calculating the vector space Euclidean distance between the current segment spiral vector index and the spiral vector index of each segment in the segment priority list; determining the relative sorting position between the spiral vector index of the segment to be sorted and the two adjacent segments in the segment priority list according to the size of the Euclidean distance, ensuring that all insertion rank position calculation results remain monotonic and there is no cross-contradiction; repeating the above operation for all segments until the clear insertion rank positions of all segment spiral vector indices relative to the existing segments in the segment priority list are determined; according to all the clear insertion rank positions obtained, all segment spiral vector indices are inserted into the corresponding positions of the segment priority list in sequence to obtain a complete monotonically increasing segment priority list.

9. The laboratory particle-gas collaborative detection method based on aerosol adsorption mechanism according to claim 8, characterized in that: Monotonically increasing is defined as follows: the Euclidean distance between the spiral vector indices of any two adjacent segments in the segment priority list shows a gradually increasing trend.

Images