A method and system for femtosecond micro-plasma molecular fingerprinting

Abstract

The application discloses a femtosecond microplasma molecular fingerprint identification method and system, utilizes nonlinear interference of three high-intensity femtosecond laser filaments to establish a transient plasma grating channel, generates a local superstrong electric field in an antinode area, and through a spatial constraint effect of the grating, makes fragments experience a chain evolution process of a coulomb explosion and continuous collision in the high-density plasma channel. Different samples have physical differences due to initial molecular topological structures, and specific path non-elastic collisions, charge transfer and energy transfer interactions between fragment groups occur when the samples are dissociated, and are reflected in dynamic evolution of characteristic spectra. By synchronously collecting emission spectra at different delay times, a wavelength-intensity-delay time three-dimensional dynamic molecular fingerprint spectrum is constructed, interactions between molecules and dissociation competition are mapped, after dynamic correlation parameters are extracted, time slice characteristics are analyzed based on a machine learning model, an overall evolution process is inversed, and nondestructive reconstruction of a molecular spatial topological configuration is realized.

Description

Technical Field

[0001] This application relates to the field of spectral analysis technology, and more specifically, to a femtosecond micro-region plasma molecular fingerprinting method and system.

[0002] This invention belongs to the field of laser molecular spectroscopy analysis and precise detection of micro-area material structure. Specifically, it relates to a dynamic molecular configuration analysis method based on femtosecond plasma grating confinement. This method breaks through the dependence of traditional spectroscopy on steady-state characteristics, establishing a rigorous spatiotemporal mapping between the underlying atomic physics mechanisms such as tunneling ionization and Coulomb explosion induced by extreme local light fields and the chemical kinetic paths of macromolecular skeleton dissociation and secondary fragment decomposition. By capturing the full-cycle radiation events of non-equilibrium evolution within the plasma channel, a high-dimensional dynamic molecular fingerprint spectrum is constructed, thereby achieving high-fidelity, lossless reconstruction of molecular spatial topology and isomers in complex systems. Background Technology

[0003] In the field of material composition detection, inductively coupled plasma atomic emission spectrometry (ICP-AES) holds a crucial position due to its elemental analysis capabilities. However, when this technology is applied to the microstructural analysis of complex systems, its physical mechanisms face fundamental bottlenecks. Existing molecular fingerprinting methods typically rely on two-dimensional spectral data descriptions, simplifying spectral data into a static mapping between wavelength and light intensity. This detection mode essentially involves time integration of the material's evolution in a non-equilibrium state, implicitly assuming that discriminative information primarily originates from elemental characteristics at a single time cross-section or in a time-averaged sense. However, under strong field excitation conditions, the ionization, dissociation, and radiation processes of molecular systems exhibit significant temporal evolution characteristics. Differences in molecular spatial configuration can lead to specific differentiation in the spectral time domain by modulating fragment generation paths and collision relaxation phases. Traditional time integration operations forcibly annihilate this temporal differentiation behavior, making it impossible to effectively distinguish substances with the same elemental composition but different configurations using existing methods.

[0004] More importantly, the physical mechanism of traditional laser-induced breakdown spectroscopy (LIBS) has inherent limitations. When a single femtosecond laser beam propagates in air, the Kerr self-focusing effect and the plasma defocusing effect reach a dynamic equilibrium. This power density is insufficient to directly counteract the internal field strength of the molecule, resulting in incomplete ionization and fragmentation. Simultaneously, the unconstrained plasma rapidly expands and cools outward, interrupting the synergistic effect of multiple physical fields, including optical field induction, ion collisions, and plasma excitation. This break in the synergistic chain causes fragments carrying the original molecular topological information to lose their configurational imprint during thermal motion, further exacerbating the loss of front-end physical information.

[0005] This lack of physical information is particularly critical in complex mixture systems, incomplete feature samples, or small sample analysis scenarios. Because front-end physical detection methods cannot capture complete dissociation dynamics information, traditional analysis methods are forced to rely heavily on complex back-end data processing algorithms to reconstruct and guess the incomplete data using purely mathematical methods. This paradigm of "insufficient front-end physical feature extraction followed by forced reconstruction through data processing" easily introduces mathematical artifacts, significantly weakening the direct, in-situ physical optical detection advantages that plasma spectroscopy should possess. Therefore, a technological breakthrough is urgently needed to capture concurrent radiation events throughout the entire non-equilibrium evolution process by increasing the local field strength and applying spatial constraints. This would establish an intrinsic dynamic mapping between the underlying physical mechanisms such as tunneling ionization, Coulomb explosions, and continuous collisions induced by extreme light fields, and chemical evolution paths such as molecular skeleton dissociation and deep fragment decomposition. This would ensure high-dimensional information extraction from the underlying physical mechanisms, thereby restoring the intrinsic advantages of plasma spectroscopy in the analysis of micro-region structures in complex systems. Summary of the Invention

[0006] The purpose of this application is to overcome the shortcomings of existing technologies by providing a femtosecond micro-region plasma molecular fingerprinting method and system. By capturing and analyzing the full-time-domain dissociation dynamics information of molecules under strong field, it breaks through the limitations of traditional static spectroscopy technology, can distinguish isomers without pre-separation, and can realize the simultaneous identification of multiple components in complex mixtures.

[0007] The objective of this application is achieved through the following technical solution: Firstly, this application proposes a femtosecond micro-region plasma molecular fingerprinting method, comprising: Multiple femtosecond laser beams are focused and interfered in a micro-region of the sample to form a plasma grating. A localized ultra-strong electric field is generated in the antinode region of the plasma grating, thereby breaking down the chemical bonds in the sample in a non-thermal manner. By utilizing the spatial confinement effect of a plasma grating, a high-density plasma channel is formed and maintained in the antinode region. The generated fragments undergo a chain evolution process within the channel and radiate characteristic spectra associated with the molecular spatial configuration. Simultaneously collect emission spectra emanating from the plasma channel at different delay times to construct a three-dimensional dynamic molecular fingerprint spectrum encoding fragment evolution dynamics based on wavelength, intensity, and delay time. From the three-dimensional dynamic molecular fingerprint spectrum, dynamic correlation parameters of multiple comprehensive evolution laws, including intensity envelope evolution, peak time difference of cross-element spectral peaks and linewidth evolution, are extracted. The PSO-DMTSNet machine learning model is used to analyze dynamic correlation parameters in order to reconstruct and identify the spatial topological configuration of molecules in the sample to be tested.

[0008] In one possible implementation, the step of using multiple femtosecond laser beams to focus and interfere in a micro-region of the sample to form a plasma grating includes: A femtosecond laser system with a center wavelength of 800 nm, a pulse width of 80 fs, and a single pulse energy of 2 mJ was used. Based on the femtosecond laser system, multiple femtosecond laser beams are focused onto the surface of the sample to be tested. During the propagation of the femtosecond laser in the air, the dynamic balance between the Kerr self-focusing effect and the plasma defocusing effect generates a stable femtosecond laser filament, forming the plasma single-filament structure required for filament-induced breakdown spectroscopy (FIBS). Split at least two femtosecond laser beams at a spatial angle When two or more such laser filaments intersect, their interference produces a periodicity (period 1 / 2). The gain fringes of the ) are significantly enhanced when the nonlinear self-focusing and plasma defocusing are balanced, thereby creating a transient plasma grating channel; in , The wavelength of the laser; The plasma grating, through coherent superposition, generates a periodic electric field enhancement in the antinode region, thereby increasing the local light intensity to [a certain level]. Quantity, of which The number of coherent beams. This represents the saturation intensity of a single filament.

[0009] In one possible implementation, when two femtosecond laser beams are intersected and interfered, a two-dimensional transient plasma grating structure is formed, which is labeled as a plasma grating-induced breakdown spectrum (GIBS) structure. When a nonlinear interference of three high-intensity femtosecond laser filaments is used to establish a plasma grating channel, a two-dimensional grating structure is formed, denoted as a three-filament coupled transient interference spectroscopy (TIBS) structure. This configuration significantly enhances the peak laser intensity and electron density, resulting in spectral characteristics that maintain molecular-level specificity while providing the sensitivity required for accurate material characterization, particularly in complex electrochemical systems where traditional methods often fall short.

[0010] In one possible implementation, the generated fragments undergo a chain-like evolution process within the channel, including: Under the influence of a localized ultra-strong electric field, molecular bonding electrons are instantaneously extracted through the quantum tunneling effect, triggering avalanche ionization and forming a high-density electron cluster. Extreme charge imbalance triggers a violent Coulomb explosion, which explosively dissociates the molecular structure into atomic ions and molecular fragments. The generation path of the fragments is related to the spatial topology and bond energy distribution of the original molecule. The dissociated fragments, confined in the submicron space of the plasma channel, undergo continuous inelastic collisions with high-density electrons and are excited, subsequently emitting characteristic spectra through radiative transitions.

[0011] In one possible implementation, the steps of constructing a wavelength-intensity-delay time three-dimensional dynamic molecular fingerprint spectrum encoding fragment evolution dynamics include: An enhanced charge-coupled device (ICCD) detector with a time-domain gate width ≤ 1 ns was used to acquire spectral signals; Construct a three-dimensional matrix of wavelength-intensity-delay time As a three-dimensional dynamic molecular fingerprint spectrum, in which To delay time, For wavelength, The intensity of light is considered. Different samples (including complex systems of organic and inorganic substances) have physical differences in their initial molecular topology. When they undergo dissociation under the high-density confinement of a plasma channel, the fragment groups derived from their complex structures will interact through inelastic collisions, charge transfer, and energy transfer along specific paths. Even in systems with multi-component doping or complex mixtures, the dynamic coupling process between high-energy particles and free radicals generated by the dissociation of different molecular groups is still strictly modulated by the steric hindrance of the original matrix, the distribution of chemical bond energy, and the coordination environment. This microscopic collision network driven by underlying interaction forces directly changes the excitation probability and de-excitation sequence of related particles, and is objectively reflected in the dynamic evolution of the emitted characteristic spectra.

[0012] This three-dimensional dynamic molecular fingerprint spectrum not only carries information on fragment electronic state transition energy levels in the wavelength dimension, records the non-equilibrium particle number distribution in the intensity dimension, and encodes the reaction phase in the time delay dimension, but also accurately maps the deep elemental coordination structure and spatial topological relationships within complex doped systems. In other words, molecules not only carry their own distinctive topological fingerprint information, but the interactions, dissociation competition, and energy transfer relationships between different molecules (such as alkanes, aromatics, colloids, and asphaltenes) in the original complex system are constantly dynamic. By capturing this high-dimensional dynamic intermolecular coupling and dissociation relationship, we can accurately invert and infer the underlying characteristics from the most fundamental physical mechanisms.

[0013] In one possible implementation, the characteristic spectrum includes the hydrogen atom line H 656.3 nm, the carbon atom line C 247.8 nm, the oxygen atom line O 777.2 nm, the nitrogen atom line N 746.8 nm, and the cyano radical band CN 388.3 nm.

[0014] In one possible implementation, a dynamically correlated parameter is used to characterize the triple coupling mapping relationship between molecular configuration and spectral evolution, including: The dynamic relationship between the evolution of spectral line intensity and linewidth of a specific element; The set of time differences between the peak occurrence times of characteristic spectral peaks of different elements; Global characteristics of the high-dimensional correlation matrix of the full-band spectrum co-evolving over time.

[0015] In one possible implementation, the machine learning model is the PSO-DMTSNet dynamic masked molecular fingerprint spectrum temporal modeling and recognition model. The PSO-DMTSNet model employs a dynamic mask consistency learning strategy, combining short-term spectral feature extraction with long-term evolutionary dependency modeling, and incorporating a dual-channel mechanism of feature attention and temporal attention. This algorithm captures short-term features of a single slice through one-dimensional convolution, tracks long-term temporal evolution using a BiLSTM network, assigns adaptive weights to different spectral bands using feature attention to identify intrinsic correlations between spectra, and focuses on key time points of dissociation and de-excitation using temporal attention. By comprehensively analyzing feature information on time slices, the model accurately inverts the overall physical process of molecular dissociation evolution using slice information, extracting multi-timescale features from a three-dimensional dynamic molecular fingerprint spectrum. The model utilizes the Particle Swarm Optimization (PSO) algorithm to automatically optimize network hyperparameters in order to analyze intensity curvature abrupt change points and cross-band peak time difference configuration sensitive parameters in the three-dimensional dynamic molecular fingerprint spectrum.

[0016] Secondly, this application proposes a femtosecond micro-region plasma molecular fingerprinting system for performing the method as described in the first aspect, including: The femtosecond laser generator module is used to generate multiple femtosecond laser beams and control their interference to form a plasma grating; The sample micro-area interaction module is used to hold the sample to be tested and provide a micro-area breakdown environment under the constraint of a plasma grating. The time-domain resolved spectral acquisition module, including an ICCD detector and a spectrometer, is used to synchronously acquire emission spectra at different delay times; The data processing and recognition module has a built-in memory and processor. The processor is configured to execute instructions to construct a three-dimensional dynamic molecular fingerprint spectrum, extract dynamic correlation parameters, and run a machine learning model to output molecular spatial topological configuration recognition results.

[0017] The main solution and its various further alternatives described above can be freely combined to form multiple solutions, all of which are solutions that can be adopted and are claimed in this application; furthermore, the (non-conflicting alternatives) can also be freely combined with each other and with other alternatives. Those skilled in the art, after understanding the solution of this application, will realize from the prior art and common general knowledge that there are many combinations, all of which are technical solutions to be protected in this application, and will not be exhaustively listed here.

[0018] This application discloses a femtosecond micro-region plasma molecular fingerprinting method and system. It utilizes the nonlinear interference of three high-intensity femtosecond laser filaments to establish a plasma grating channel. This configuration can create a transient plasma grating structure, generating a localized ultra-strong electric field in the antinode region, significantly enhancing the peak laser intensity and electron density, thereby creating the necessary conditions for achieving efficient chemical bond breaking through tunneling ionization and Coulomb explosion processes. Utilizing the spatial confinement effect of the grating, within the formed high-density plasma channel, the fragments undergo a chain evolution process including Coulomb explosions and continuous collisions. Different samples (including complex systems of organic and inorganic substances) exhibit physical differences in their initial molecular topology. During dissociation under the high-density confinement of a plasma channel, fragments derived from these complex structures undergo inelastic collisions, charge transfer, and energy transfer along specific paths, objectively reflected in the dynamic evolution of the emitted characteristic spectra. Simultaneous acquisition of emission spectra at different delay times constructs a three-dimensional dynamic molecular fingerprint spectrum of wavelength, intensity, and delay time, accurately mapping the interactions and dissociation competition between molecules within the complex system. Dynamic correlation parameters are extracted. Based on a machine learning model, the core logic of short-term, long-term, and multi-dimensional attention mechanisms is used to analyze time-slice features, inverting the overall evolution process using slice information to achieve non-destructive reconstruction and identification of molecular spatial topological configurations. By capturing this high-dimensional dynamic intermolecular coupling and dissociation relationship, isomers can be distinguished without pre-separation, and the simultaneous and accurate identification of multiple components in complex mixtures can be achieved. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A schematic flowchart of a femtosecond micro-region plasma molecular fingerprinting method proposed in an embodiment of this application is shown.

[0021] Figure 2 The core flowchart of the femtosecond micro-region plasma molecular fingerprinting technology proposed in this application embodiment is shown.

[0022] Figure 3 A schematic diagram of the microscopic path of the dynamic chain evolution process proposed in the embodiments of this application is shown.

[0023] Figure 4 A schematic diagram of the TIBS structure for forming a two-dimensional plasma grating by interferometry of three femtosecond laser beams, as proposed in an embodiment of this application, is shown.

[0024] Figure 5 A schematic diagram of the dynamic chain process of "tunneling dissociation-Coulomb explosion-continuous collision excitation-radiative de-excitation" is shown. Detailed Implementation

[0025] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0026] Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] To address the problems of existing technologies, this application proposes a femtosecond micro-region plasma molecular fingerprinting method and system. It utilizes multi-beam femtosecond laser interference to form a spatially periodic plasma grating, overcoming the single-filament clamping effect and generating a localized ultra-strong electric field in the micro-region. This ultra-strong electric field instantaneously strips bound electrons from the molecule through quantum tunneling, non-thermally breaking chemical bonds and triggering avalanche ionization. This leads to a violent Coulomb explosion in the molecular skeleton due to instantaneous charge imbalance, achieving explosive dissociation carrying original configuration information. Subsequently, the sub-micron-level spatial confinement of the grating maintains a high-density plasma environment, driving the dissociated fragments to continuously collide with high-density electrons within the plasma channel inelastically. This maintains a long-term dynamic chain evolution process of "tunneling dissociation-Coulomb explosion-continuous collision excitation-radiative de-excitation," causing atomic and molecular fragments to radiate configuration-specific characteristic spectra.

[0028] Figure 3This illustration shows a microscopic path diagram of the dynamic chain evolution process proposed in this application. Taking the stimulated reaction of polypropylene (PP) plastic as an example, the initial PP macromolecular chain undergoes random C-C bond breakage under the action of a strong field (corresponding to "tunneling dissociation" and "Coulomb explosion"), generating a series of free radical fragments carrying information about the original chain structure. These fragments are not static, but under the constraint of a high-density plasma channel, they continuously undergo multiple competing pathways such as dissociation, 1,2-hydrogen transfer, and oxidation / nitridation reactions with O2 / N2 in the environment, dynamically evolving to generate final small molecule products such as H2CO, CO, and HCN. This complete reaction network from macromolecular dissociation to the generation of small molecule fragments is the physicochemical basis for the fragments' "continuous collision excitation" and eventual "radiative de-excitation," emitting configuration-specific characteristic spectra such as CH, C2, and CN.

[0029] By synchronously acquiring time-resolved emission spectra, a three-dimensional dynamic molecular fingerprint spectrum—wavelength-intensity-delay time—can be constructed, unfolding molecular dissociation dynamics in the time dimension. Finally, by combining machine learning to analyze the correlation parameters in this spectrum, such as the spectral line intensity envelope, peak timing difference, and dynamic evolution of linewidth, non-destructive reconstruction and accurate identification of the molecular spatial topology in complex mixtures can be achieved. This invention effectively avoids the limitation of traditional static spectral detection, which annihilates dynamic information due to time integration. It can distinguish isomers and other substances with different configurations without pre-separation, and can be extended to high-precision identification of molecular spatial configurations in various scenarios such as environmental pollutant monitoring, biological tissue diagnosis, and new material analysis.

[0030] Please refer to Figure 1 , Figure 1 This paper presents a schematic flowchart of a femtosecond micro-region plasma molecular fingerprinting method according to an embodiment of this application, including: Step S1: Multiple femtosecond laser beams are focused and interfered in a micro-region of the sample to be tested to form a plasma grating. A local ultra-strong electric field is generated in the antinode region of the plasma grating, thereby breaking down the chemical bonds in the sample in a non-thermal manner.

[0031] To achieve efficient and controllable non-thermal breakdown of chemical bonds in samples by constructing a femtosecond-plasma grating synergistic system, a spatially periodic plasma grating is formed by interfering multiple femtosecond laser beams in a micro-region of the sample. This overcomes the inherent power clamping effect when a single femtosecond laser forms a single plasma filament, which limits the local intensity at the filament core to approximately [value missing]. This is insufficient for effectively manipulating molecules. By utilizing the coherent superposition principle of multi-beam interference, a periodic electric field enhancement is generated in the antinode region of the grating, significantly increasing the local light intensity to over [amount missing]. This generates a localized, ultra-strong electric field on the order of 2, equivalent to the Coulomb field inside the molecule. This ultra-strong electric field can directly distort the potential barrier inside the molecule, and through the quantum tunneling effect, instantaneously strip the bound electrons of the chemical bonds within a femtosecond timescale, fundamentally and non-thermally breaking the chemical bonds.

[0032] The steps of using multiple femtosecond laser beams to focus and interfere in a micro-region of the sample to form a plasma grating include: A femtosecond laser system with a center wavelength of 800 nm, a pulse width of 80 fs, and a single pulse energy of 2 mJ was used. Based on the femtosecond laser system, multiple femtosecond laser beams are focused onto the surface of the sample to be tested, and a self-focusing effect occurs in the air medium. When the light intensity reaches the multiphoton ionization threshold, plasma is formed. A dynamic balance is achieved between the self-focusing effect and the plasma defocusing effect to form a plasma single-light filament; Split at least two femtosecond laser beams at a spatial angle Incident and interference occur, forming a period of The plasma grating, in which , The wavelength of the laser; The plasma grating, through coherent superposition, generates a periodic electric field enhancement in the antinode region, thereby increasing the local light intensity to [a certain level]. Quantity, of which The number of coherent beams. The saturation intensity of a single filament is greater than the local intensity. .

[0033] The efficient and stable breakdown of the sample by the plasma grating formed by femtosecond laser interferometry utilizes a femtosecond laser system with a center wavelength of 800 nm, a repetition rate of 1 kHz, and a pulse width of 80 fs, with a single pulse energy of 2.0 mJ. The focused beam is incident on the sample under test in air (with a nonlinear Kerr coefficient n² = 3.2 × 10⁻⁶). -19 (cm² / W) exhibits a self-focusing effect, its focal length satisfy: (1) When the light intensity reaches the multiphoton ionization threshold At this time, gas molecules ionize to form plasma. The plasma induces a defocusing effect, causing a change in its refractive index. As shown in equation 2, where The plasma frequency is [value missing]. The dynamic balance between self-focusing and plasma defocusing forms a stable plasma single filament with a diameter of approximately 100 μm. Due to the power clamping effect, the light intensity in the filament core region is limited to [value missing]. Electron density saturates This results in the field strength and electron density of a single filament being unable to meet the ultra-high field strength requirements for molecular quantum state manipulation. ).

[0034] (2) To overcome the clamping effect, this invention generates multiple beams (two beams in the basic model and three beams in the optimized model) with a spatial angle of [missing information]. The femtosecond filaments, with multiple filaments interfering in overlapping regions, form a spatially periodically modulated plasma grating. Grating period. By wavelength and included angle The decision is as shown in Equation 3.

[0035] (3) This grating reconstructs the spatial light field distribution through the principle of coherent superposition. (Equation 4) generates a periodic electric field enhancement in the antinode region, increasing the local light intensity to [value missing]. Magnitude ( ), corresponding electric field strength As shown in Equation 5. This enhancement mechanism improves ionization efficiency by 1-2 orders of magnitude, significantly exceeding the single-filament clamping limit in peak intensity, and achieving an electron density of... This provides the necessary ultra-strong local field for the effective breaking of chemical bonds. In particular, the two-dimensional grating structure (TIBS) constructed using three lasers can further increase the intensity of characteristic spectral signals by about 2 times, improve detection sensitivity by about 33%, and extend plasma lifetime by about 20% compared to the two-laser GIBS.

[0036] (4) (5) The core feature of this plasma grating structure lies in its exceptional stability. By precisely controlling the grating period, the plasma density and electron density are effectively enhanced and stably maintained, ensuring that the excitation process of matter is in a highly controllable and stable state. The scale-like plasma grating structure formed by multiple femtosecond filaments not only provides the necessary high power density guarantee for reliable chemical bond breaking, but its spatial periodic constraint also significantly reduces the random fluctuations in plasma evolution.

[0037] When two femtosecond laser beams are intersected to generate periodic gain fringes, a two-dimensional transient plasma grating structure is formed, which is labeled as a plasma grating-induced breakdown spectrum (GIBS) structure. When nonlinear interference is generated using three high-intensity femtosecond laser filaments, a transient plasma grating channel is established, which is labeled as the TIBS structure of the three-filament coupled transient interference spectroscopy method.

[0038] Depending on the number of interferometric laser beams used, plasmonic grating structures with different performance characteristics can be formed. When two femtosecond lasers are used for transient interferometry, the constructed two-dimensional plasmonic grating structure is labeled as plasmonic grating-induced breakdown spectrum (GIBS). This structure can generate a field strength exceeding [a certain value] in the antinode region. The localized ultra-strong field generates a plasma characteristic spectral signal intensity that is 3 times stronger than that of the traditional filament-induced breakdown spectroscopy (FIBS) structure, and the plasma lifetime is extended by 2 times. To further improve performance, three high-intensity femtosecond laser filaments are used to establish a channel through nonlinear interference. The resulting two-dimensional grating structure is labeled as three-filament coupled transient interferometry (TIBS). This three-filament coupled transient interferometry structure achieves superior overall performance, with a plasma signal intensity 4 times stronger than that of FIBS; compared with the two-beam plasma grating-induced breakdown spectroscopy (GIBS) structure, its detection sensitivity is further improved by about 33%, while the plasma lifetime is extended by about 20%. Especially in complex electrochemical systems, traditional methods are often inadequate, while this technology bridges the gap between traditional elemental analysis and intrinsic molecular-level detection capabilities.

[0039] Figure 4 This illustration shows a schematic diagram of a TIBS structure for forming a two-dimensional plasmonic grating using three femtosecond lasers, as proposed in this application. The core optical path is as follows: A femtosecond laser pulse from an fs laser is first guided by a mirror, then sequentially split by two beam splitters to generate three independent laser beams with controllable intensity. Each of these three beams is equipped with an independently adjustable displacement platform to precisely control the optical path difference and spatial angle between the beams, which is crucial for forming a stable, periodic two-dimensional interference pattern. The three beams are finally converged at the same point on the sample stage by a focusing lens, interfering in the micro-region of the sample to form a two-dimensional distributed plasmonic grating structure. This grating, through the local field strength multiplication effect in the antinode region, provides a super-strong and spatially confined excitation environment for subsequent non-thermal breakdown and fragment evolution. Simultaneously, the device is synchronously triggered by a photodetector and uses an ICCD spectrometer to acquire the time-resolved spectral signal generated by the TIBS structure, transmitting it to a PC for subsequent three-dimensional dynamic molecular fingerprint construction and intelligent analysis.

[0040] Step S2: Utilizing the spatial confinement effect of the plasma grating, a high-density plasma channel is formed and maintained in the antinode region. The generated fragments undergo a chain evolution process within the channel and radiate characteristic spectra associated with the molecular spatial configuration. The core mechanism of step S2 lies in utilizing the extreme spatial confinement characteristics of the plasma grating formed in step S1. This grating confines the excitation region within a submicron-scale antinode channel, thereby forming and maintaining a high-electron-density plasma environment for a long period. In this high-density confinement environment, the sample broken down by the localized ultra-strong electric field undergoes an ordered and rapid chain-like physical evolution process: First, the strong electric field instantaneously strips the molecular bonded electrons through the quantum tunneling effect, triggering avalanche ionization; then, due to the instantaneous imbalance of charge, the molecular skeleton undergoes a violent Coulomb explosion, explosively dissociating into initial fragments carrying the original molecular spatial topological information; subsequently, during and after the femtosecond laser pulse, these fragments cannot diffuse freely under the confinement of the grating channel, and undergo continuous and repeated inelastic collisions with high-density electrons. This "continuous collision excitation" process continuously excites the fragments to high-energy states, which are then de-excited through radiative transitions. The entire dynamic chain of "tunneling ionization → Coulomb explosion → collision excitation → radiative de-excitation" causes the generated atomic and molecular fragments to radiate characteristic spectra throughout their evolution. The temporal characteristics of these spectra, such as intensity envelope and appearance sequence, are directly modulated by the initial molecular configuration, thus radiating characteristic spectra related to the molecular spatial configuration.

[0041] The generated fragments undergo a chain-like evolutionary process within the channel, including: Under the influence of a localized ultra-strong electric field, molecular bonding electrons are instantaneously extracted through the quantum tunneling effect, triggering avalanche ionization and forming a high-density electron cluster. Extreme charge imbalance triggers a violent Coulomb explosion, which explosively dissociates the molecular structure into atomic ions and molecular fragments. The generation path of the fragments is related to the spatial topology and bond energy distribution of the original molecule. The dissociated fragments, confined in the submicron space of the plasma channel, undergo continuous inelastic collisions with high-density electrons and are excited, subsequently emitting characteristic spectra through radiative transitions.

[0042] Peak intensity generated in the antinode region of the plasma grating Under the localized ultra-strong field, the interaction between laser and matter completely departs from the thermodynamic equilibrium and incomplete dissociation process of traditional spectroscopy. This extreme light field triggers a series of ordered and tightly connected ultrafast dynamic chains dominated by underlying quantum physics mechanisms. Based on the evolution of the physical driving force, this dynamic process of "tunneling dissociation - Coulomb explosion - continuous collision excitation - radiative de-excitation" can be divided into the following three stages: Field-induced tunneling ionization and Coulomb explosion dissociation stage: This stage begins with the direct physical intervention of an extreme micro-area optical field on the Coulomb barrier inside the polymer molecule. Under the action of a powerful oscillating laser electric field, the barrier faced by the bound electrons on the molecule's periphery is extremely distorted, and electrons are directly and instantaneously extracted within an fs ultrafast timescale through quantum tunneling. Since the essence of a chemical bond is the sharing of electron clouds, the non-thermal extraction of electrons directly and fundamentally destroys the electronic basis of the chemical bond, triggering primary field-induced dissociation; the released seed electrons are accelerated to collide with neighboring atoms under the drive of the strong optical field, triggering primary avalanche ionization. In the instant when a large number of electrons are removed rapidly, the molecular skeleton, which has lost its charge balance, generates an unsustainable huge Coulomb repulsive force, directly triggering a violent Coulomb explosion. This physical process is much faster than molecular vibrational relaxation (on the order of ps), explosively dissociating the original molecular structure into an initial fragment group carrying local configuration information. This nonthermal dissociation, dominated by Coulomb forces, strictly depends on the spatial topology and bond energy distribution of the original molecule: for example, the high-bond-energy, rigid benzene ring skeleton in polystyrene (PS) tends to break at the weaker bonds at the edges during a Coulomb explosion, preserving large-volume benzene ring fragment ions (such as phenyl cations). ), and release benzyl radicals ( The flexible aliphatic long chains of polyethylene (PE) are prone to statistical multi-point random breakage under Coulomb repulsion, mainly generating alkyl chain cations (such as...). or ) and ethyl radical ( Polyethylene terephthalate (PET), which contains highly polar ester bonds (-COO-), will preferentially break at the carbonyl site under the induction of a physical field, generating a benzoyl cation. and ethoxy radicals .

[0043] Collision Avalanche and Fragmentation Stage under Extreme Constraints: With the continuation of the pulse and the completion of primary dissociation, the evolving system enters a deep decomposition phase driven by the synergistic effects of residual strong light fields, high-density free electron flows, and ion swarms. The submicron-level spatial confinement effect of the plasma grating plays a decisive role here, tightly locking the initially generated macromolecular fragment ions and highly reactive free radicals within an extremely small, high-density space, significantly enhancing the proximity effect and collision cross-section between particles. Under this confinement environment, high-energy free electrons continuously bombard macromolecular fragments, maintaining and amplifying the avalanche ionization process within the channel; simultaneously, the large amount of kinetic energy transferred by heavy ion collisions triggers secondary breakage of weak bonds. This secondary transformation driven by physical collisions directly leads to the deep decomposition of the initial macroscopic large fragments. For example, the benzyl free radicals generated during primary dissociation... After undergoing intense high-frequency physical collisions, its methylene CH bond is further broken, generating a smaller benzyl radical. and ; benzyl cation It may also be lost in a collision. Transform into The core physical result of this stage is that the macromolecules within the system are completely pulverized, generating a diverse array of extremely small secondary fragments, containing a large number of small molecule free radicals (such as methyl free radicals). hydroxyl radicals cyano radical ), simple diatomic molecules (such as , The extremely high collision frequency maintained by the grating, along with the significant emergence of atomic-level fragments (C, H, O, N, etc.), is the key physical engine that accelerates this deep atomization process.

[0044] Continuous collisional excitation and characteristic radiation de-excitation stage: After the femtosecond laser pulse completely ends, the direct driving force of the optical field on the debris disappears, and the evolution is completely dominated by the collisional dynamics inside the plasma channel, entering a radiation development period lasting from nanoseconds to microseconds. Due to the high-density confinement of the grating channel significantly suppressing the free diffusion of particles, high-energy electrons, various ions, and secondary debris groups form a continuous, intertwined dynamic physical chain reaction of "dissociation-collision-excitation-emission" within the channel. The continuous bombardment of high-energy electrons not only further promotes atomization but also excites the debris to specific high-energy states through inelastic collisions; simultaneously, secondary debris in the excited state (such as...) or Energy is efficiently transferred to neighboring ground-state particles through collisions. Newly formed atoms may also collide to form transient excited complexes (such as...). Subsequently, these particles in specific high-energy states (such as...) , , , , , , The fragments release characteristic photons through radiative transitions. Since the energy transfer, excitation probability, and radiative relaxation in this stage are entirely controlled by the collision dynamics between fragments, the spectra of different elements exhibit strong configuration-sensitive temporal evolution characteristics: for example, the rapid rising edge of the hydrogen atom line (H 656.3 nm) physically maps the rate at which hydrogen fragments are continuously stripped from the parent macromolecule and excited; while the spectral envelope evolution of nitrogen-containing molecular bands is limited by the release efficiency of precursor fragments and the probability of recombination reactions. The stable micro-region confinement environment of the plasma grating allows fragments to fully exhibit their collision dynamics within a sufficiently long time window, thus perfectly amplifying and encoding the path differentiation caused by the molecular micro-topology (such as the spatial steric hindrance difference between rigid rings and flexible chains) into macroscopically measurable spectral temporal and linear dynamic evolution characteristics.

[0045] Figure 5 The diagram illustrates a dynamic chain process of "tunneling dissociation - Coulomb explosion - continuous collisional excitation - radiative de-excitation," which can be clearly divided into three stages based on time evolution and physical mechanisms: The first stage (around 0 ns) is ultrafast excitation and primary dissociation. Correspondingly, in the sample on the left side of the diagram, under the synergistic effect of a femtosecond-plasma grating, the molecular structure is non-thermally broken down by tunneling ionization under a localized ultra-strong electric field, and then explosively dissociates due to a Coulomb explosion, generating initial highly active fragments and plasma; the second stage (around 0 ns)... The first stage (from ns to hundreds of ns) represents fragment collision evolution and secondary reactions. Corresponding to the particle swarm in the middle of the figure, which fades in color over time, it shows that the initial fragments undergo continuous decomposition, collision, excitation, and complex reactions (such as oxidation and recombination) with environmental molecules within a grating-confined high-density plasma channel, dynamically evolving into simpler molecular or atomic species. The third stage (covering the entire time domain) represents characteristic radiation and spectral fingerprint generation. Corresponding to the characteristic spectral peaks of C, H, and O that change over time at the top of the figure, it indicates that the excited atomic and molecular fragments in the above evolution process emit characteristic spectra with specific time evolution patterns through radiation de-excitation. These spectral lines that change dynamically over time together constitute a three-dimensional dynamic fingerprint for identifying the original molecular configuration.

[0046] Furthermore, the evolution process after the pulse ends includes: the duration of the evolution process after the pulse action is dynamically extended according to the excited state lifetime of different material systems and different fragments and their collision history in the plasma. Under extreme constraints, it can continue to the millisecond timescale. The wavelength shift, linear profile and intensity decay behavior of its radiation spectrum condense the local chemical environment and initial molecular structure information in multiple dimensions.

[0047] The characteristic spectra include the hydrogen atom line H 656.3 nm, the carbon atom line C 247.8 nm, the oxygen atom line O 777.2 nm, the nitrogen atom line N 746.8 nm, and the cyano radical band CN 388.3 nm.

[0048] Step S3: Simultaneously collect emission spectra radiated from the plasma channel at different delay times to construct a three-dimensional dynamic molecular fingerprint spectrum of wavelength-intensity-delay time that encodes fragment evolution dynamics.

[0049] Step S3, the construction and dynamic correlation decoupling of time-resolved molecular fingerprint spectrum, is fundamentally about overcoming the two-dimensional limitations of traditional static spectroscopy. It involves fully "unfolding" the ultrafast dissociation dynamics of molecules in the time dimension and encoding it into a resolvable data structure. Using an enhanced charge-coupled device (CCD) detector with a time gate width of 1 nanosecond or less, emission spectra emitted from the plasma channel at different delay times can be acquired simultaneously in 0.1 ns steps. By systematically acquiring a series of instantaneous spectral "snapshots" from the moment of laser breakdown to microseconds or even milliseconds later, the entire temporal evolution trajectory of the molecule—from steady-state structural collapse, through Coulomb explosion dissociation, to fragment collision excitation and recombination radiation—is fully recorded. Finally, these time-series acquired two-dimensional spectra are integrated to construct a time-resolved molecular fingerprint spectrum based on wavelength (…). ), light intensity ( ) and delay time ( () is a three-dimensional data matrix with three orthogonal dimensions, i.e. Three-dimensional dynamic molecular fingerprint spectrum. In its physical essence, this fingerprint spectrum is not a simple superposition of spectra, but rather carries information about the specific electronic transition energy levels of fragments in the wavelength dimension, reflects the particle population and abundance of various fragments in the non-equilibrium state in the intensity dimension, and precisely encodes the reaction paths and phase differences of different fragment generation, excitation and radiative de-excitation in the time delay dimension. Thus, it realizes the information encoding and visualization of the complete dynamic chain of "tunneling ionization-Coulomb explosion-collision excitation".

[0050] The steps involved in constructing a wavelength-intensity-delay time three-dimensional dynamic molecular fingerprint spectrum encoding fragment evolution dynamics include: An enhanced charge-coupled device (ICCD) detector with a time-domain gate width ≤ 1 ns was used to acquire spectral signals; Construct a three-dimensional matrix of wavelength-intensity-delay time As a three-dimensional dynamic molecular fingerprint spectrum, in which To delay time, For wavelength, The light intensity is represented by the three-dimensional dynamic molecular fingerprint spectrum, which carries information on the energy levels of fragment electronic state transitions in the wavelength dimension, records the distribution of non-equilibrium particle numbers and fragment abundance in the intensity dimension, and encodes the reaction path and de-excitation evolution phase in the delay time dimension.

[0051] Based on the multi-stage dynamic evolution of fragments within a femtosecond plasma grating channel, this invention constructs a core identification carrier—time-domain resolved molecular fingerprinting—through high-time-resolution detection technology and data organization reconstruction. When describing plasma molecular fingerprinting and characteristic spectra, it is important to emphasize that different samples (covering complex systems including organic and inorganic materials) exhibit physical differences in their initial molecular topology. During dissociation under the high-density confinement of the plasma channel, fragment groups derived from their complex structures undergo interactions such as inelastic collisions, charge transfer, and energy transfer along specific paths. Even in systems with multi-component doping or complex mixtures, the dynamic coupling process between high-energy particles and free radicals generated by the dissociation of different molecular groups is still strictly modulated by the steric hindrance of the original matrix, the distribution of chemical bond energy, and the coordination environment. This microscopic collision network driven by underlying interaction forces directly alters the excitation probability and de-excitation sequence of related particles, objectively reflecting in the dynamic evolution of the emitted characteristic spectra. The final constructed plasma molecular fingerprint spectrum can accurately map the deep elemental coordination structure and spatial topological relationships within complex doped systems. This means that molecules not only carry their own distinctive topological fingerprint information, but the interactions, dissociation competition, and energy transfer relationships between different molecules (such as alkanes, aromatics, colloids, and asphaltenes) in the original complex system are constantly dynamic. By capturing this high-dimensional "dynamic intermolecular coupling and dissociation relationship," we can accurately invert and infer the underlying characteristics from the most fundamental physical mechanisms. The physical necessity of this construction step stems from the intrinsic irreversibility of plasma evolution and the continuity of dissociation dynamics, specifically including the following three progressive construction and decoupling cores: First, overcoming the limitations of two-dimensional static methods and constructing three-dimensional observation functions: In existing technologies, molecular fingerprint spectra are typically processed as wavelength... With light intensity For two-dimensional functions, this static representation implicitly incorporates the time dimension. This is considered an interfering factor that needs to be eliminated or integrally averaged, aiming to obtain a spectral snapshot at a fixed moment. However, under extremely strong field excitation conditions, the observable spectrum of a molecule is not a direct mapping of its static molecular structure, but rather a comprehensive physical manifestation of the molecule in dynamic processes such as tunneling ionization, Coulomb explosion, collisional excitation, and de-excitation radiation. Therefore, this invention employs a time-domain gate width... The ICCD detector continuously acquires spectral signals, elevating the time dimension to a core dimension of equal importance to wavelength and light intensity, and constructing a three-dimensional function of wavelength-intensity-delay time. : (6) From a physical modeling perspective, if we denote the dissociation evolution state of a molecular system under a strong field as... Then the instantaneous spectrum collected at this time Essentially, it is a physical mapping from the microscopic dynamic state to a macroscopically observable signal. This three-dimensional fingerprint spectrum carries information on fragment electronic state transition energy levels and local chemical environment in the wavelength dimension, records the non-equilibrium particle number distribution and fragment abundance in the intensity dimension, and encodes reaction paths and de-excitation evolution phases in the time-delay dimension, achieving a complete mapping of the plasma radiation field in the time-spectral domain. Simultaneously, this three-dimensional function extends data representation from the traditional two-dimensional plane to three-dimensional space, carrying electronic state transition energy levels in the wavelength dimension, recording the non-equilibrium particle number in the intensity dimension, and encoding reaction path phases in the time dimension.

[0052] Second, reconstructing a continuous temporal trajectory that breaks through the limitations of discrete observation: Simply acquiring multi-time-point data on hardware is insufficient to support dynamic analysis; it is essential to address the critical issue of expressing the temporal correlation of molecular evolution at the data level. In the original spectral acquisition, to obtain statistical stability, experiments typically involve multiple repeated measurements at each delayed time point, inevitably resulting in the original data being stored in an interleaved manner in the form of "time point × number of repetitions." While this discrete storage method facilitates experimental recording, it physically severs the continuous evolution process of the same molecule under the same strong field excitation, making it difficult for subsequent models to directly learn the continuous changes in molecular states. To address this, this step introduces a specialized temporal reconstruction mechanism, which strictly reorganizes the instantaneous spectra of the same molecular sample and the same experiment, acquired at different time points, according to absolute chronological order. This reconstruction process achieves a core physical transformation: seamlessly recombining the originally interleaved discrete observation fragments into a continuous temporal trajectory reflecting the same molecule over a complete evolutionary cycle.

[0053] Third, dimensionality reduction and decoupling of state evolution mapping and dynamic correlation: After 3D modeling and temporal reconstruction, the principle of using the state evolution function corresponding to the molecular dissociation path as the core of discrimination was established. Within the plasma channel, different molecular configurations determine the order of generation of characteristic fragments and the collision cross-section. These microscopic collision networks are decoupled into macroscopically analyzable correlation topologies in the reconstructed continuous trajectories. For example, the inelastic collision history of excited-state atoms is mapped to the phase covariance of specific spectral line intensities, and the spatial association and condensation of complementary fragments is a delayed locking of the intensity envelope. Within this physical framework, subsequent analysis and identification are no longer feature extraction from static spectra, but rather learned... The continuous temporal structure, under data-driven conditions, approximates the actual state evolution trajectory corresponding to the microscopic dynamic system. This fundamentally shifts the physical basis for identifying complex systems from the traditional static spectral characteristics to whether the physical dynamic evolution paths of molecular dissociation and recombination are consistent.

[0054] Step S4: Extract dynamic correlation parameters from the three-dimensional dynamic molecular fingerprint spectrum, including the evolution of intensity envelope, the peak time difference of cross-element spectral peaks, and the evolution of spectral linewidth. The analysis of the configuration-sensitive dynamic fingerprint mapping mechanism in step S4 is to analyze the wavelength-intensity-delay time three-dimensional dynamic molecular fingerprint spectrum constructed in step S3. In this process, the molecular spatial topological information encoded in the spectral temporal evolution is decoded. First, dynamic parameters characterizing the dynamic correlations need to be extracted from the three-dimensional data matrix. These mainly include: 1) the evolution curve I of the intensity of specific characteristic spectral lines with time delay and its envelope shape; 2) the peak occurrence times of characteristic radiation peaks of different element atoms or molecular fragments, and the calculation of the time difference between each pair to determine the dynamic changes in linewidth of specific spectral lines during the evolution process. These parameters do not exist in isolation, but rather form a dynamic fingerprint network through a triple physical coupling mechanism: "dynamic linkage between the intensity and linewidth of specific element spectral lines," "time difference across elemental peaks," and "global characteristics of the full-time-spectral correlation matrix." The spatial configuration differences of the original molecule modulate the fragment generation path after the Coulomb explosion and its subsequent collision relaxation phase, thus accurately mapping to the specific differentiation of macroscopically measurable parameters such as the slope of the intensity envelope, the value of the peak time difference, and the rate of linewidth evolution.

[0055] Dynamic correlation parameters are used to characterize the triple coupling mapping relationship between molecular configuration and spectral evolution, including: The dynamic relationship between the evolution of spectral line intensity and linewidth of a specific element; The set of time differences between the peak occurrence times of characteristic spectral peaks of different elements; Global characteristics of the high-dimensional correlation matrix of the full-band spectrum co-evolving over time.

[0056] Time-domain resolved molecular fingerprinting The three-dimensional matrix depth carries the dynamic fingerprint information of molecular configuration. The essence of this fingerprint spectrum is a dynamic holographic projection of the original molecule during plasma evolution, and its core value lies in the strict physical correlation of spectral characteristic parameters in the temporal evolution. These correlations do not exist in isolation, but are condensed into resolvable dynamic identifiers through the emission behavior and interactions of characteristic atomic / molecular spectral lines (such as H 656.3nm, C 247.8nm, O 777.2nm, N 746.8nm, CN 388.3nm) of C, H, O, N, and CN. The core of this mapping mechanism is that the spatial topological differences of the original molecule directly modulate the evolution path and collision cross-section of fragments within the plasma channel, leading to configuration-sensitive differentiation in the temporal characteristics of characteristic spectral line radiation, specifically manifested as the following triple physical coupling mapping relationship: 1. Dynamic linkage and steric hindrance mapping of spectral line parameters of specific elements. Intensity evolution curve of a single characteristic element spectral line (e.g., hydrogen atom H 656.3 nm). Its spectral linewidth The changes in these processes form a tight coupling. This correlation between intensity and linewidth is not random, but directly constrained by the molecular spatial structure. For example, when a molecule has a rigid fused ring configuration, the steric hindrance of the generated hydrogen fragments increases, and the collision frequency decreases, which will synchronously lead to... The rising edge is slower and the peak is delayed, and The broadening rate is suppressed. Conversely, fragments generated by flexible chain molecules (such as polyethylene) have low steric hindrance and active collisions, resulting in faster intensity evolution and significant dynamic fluctuations in linewidth. This parameter linkage between intensity and linewidth is direct evidence that the fragment relaxation path is modulated by the molecular spatial topology.

[0057] 2. Correlation encoding between inter-elemental radiation peak timing differences and chemical bond breaking sequences. The characteristic radiation peaks of different elements exhibit an ordered correlation sequence over time. The peak occurrence times of C, H, O, N, CN, etc. Constitutes the key time difference set This temporal relationship is essentially a direct mapping between the sequence of chemical bond breaking within the molecule and the fragment release kinetics; fragments released from the breaking of weak bonds (such as aliphatic CH bonds)... Its radiation peak (H) usually precedes that of strong bonds (such as aromatic C=C bonds) fragments. The radiation peak of ) (C 247.8nm), that is The peak time difference between the oxygen correlation line (O 777.2 nm) and the carbon correlation line (C 247.8 nm) of molecules containing specific functional groups (such as PET containing ester groups). The relative timing of ester bond cleavage is directly correlated; the peak time of the CN radical band (388.3 nm) may reflect the release efficiency or recombination probability of nitrogen-containing groups. Subtle differences in molecular configuration (such as variations in bond energy distribution in isomers) will precisely modulate these interelemental timing differences. This results in a dynamic gradient encoding of chemical bond breaking sequences on the spectral time axis.

[0058] 3. Configurational fingerprint and global topological response of the full-time spectral correlation matrix. The fingerprint spectrum, as a whole, constitutes a high-dimensional correlation matrix through the co-evolution patterns of its full-band spectra along the time axis. The global properties of this matrix (such as the covariance of intensity evolution in specific wavelength ranges and the distribution patterns of antisymmetric components) remain highly sensitive to the molecular three-dimensional configuration (such as chirality and functional group orientation). Differences in three-dimensional configuration are transformed into sign differences or distribution shifts of characteristic components in the matrix by influencing the spatial phase distribution of fragment wave packets and collision geometry. Simultaneously, the surrounding environmental medium (such as air) also plays a role. Characteristic spectral lines generated by the molecule (such as OH 309nm, N 746.8nm) also participate in the evolution, and their interaction modes with the target molecule's spectral lines (such as intensity ratio) also play a role. The evolution further enriches the dynamic fingerprint of the hybrid system, but the micro-region constraints of the grating and the dominance of evolutionary dynamics still allow the configurational features of the target molecule to be separated from the complex correlation network.

[0059] The logical connections between the aforementioned features collectively constitute a dynamic fingerprint network of molecular configuration. The complex temporal evolution spectrum is not an obstacle, but rather a unique dynamic fingerprint carrier that condenses the molecular configuration information before femtosecond laser excitation in the dynamic dimension. If these dynamically correlated parameters can be analyzed, subtle differences in molecular configuration can be accurately read from the spectral evolution, ultimately achieving precise identification of molecular configuration.

[0060] However, the extraction of dynamic fingerprints faces fundamental bottlenecks. A single spectral line signal is actually a convolutional superposition of multiple physical processes. The dynamic chaos caused by fragmented chain reactions exceeds the analytical limits of traditional models, and insufficient temporal resolution leads to the loss of quantum coherence information. This inevitably renders manual analysis ineffective. Dynamic noise masks the inherent constraints between parameters, temporal randomness obscures the broken bond sequence, and matrix fluctuations devour configurational identifiers. Only intelligent algorithms can identify deep correlations.

[0061] Step S5: Analyze the dynamic correlation parameters based on the machine learning model to reconstruct and identify the spatial topological configuration of molecules in the sample to be tested.

[0062] Step S5, AI Dynamic Fingerprint Intelligent Analysis and Multi-Scenario Recognition Application, focuses on the deep analysis of the dynamic correlation parameters extracted in Step S4 using a specialized machine learning model, thereby constructing a final recognition bridge from spectral dynamics to molecular structure. Specifically, an algorithm called Dynamic Masked Molecular Fingerprint Spectral Temporal Modeling and Recognition Model (PSO-DMTSNet) is employed. This model uses a dynamic mask consistency learning strategy to force the network to output consistent recognition results for incomplete temporal observation samples, enabling it to internalize the evolutionary laws of molecular dissociation dynamics rather than memorizing fixed spectral fragments. Its network structure combines one-dimensional convolution and bidirectional long short-term memory (BiLSTM) networks, supplemented by a dual-channel attention mechanism, to achieve multi-timescale feature extraction from short-term local collision events to long-term evolution. Through automatic optimization of network hyperparameters using a particle swarm optimization algorithm, this model can accurately analyze key configuration-sensitive parameters in the three-dimensional dynamic fingerprint spectrum, such as curvature abrupt changes in characteristic spectral line intensity evolution and peak temporal differences between cross-elemental spectral peaks. Ultimately, the model establishes a complete mapping from "dynamic spectral sequence → high-dimensional physical evolution characteristics → molecular spatial topology," forming a closed loop for intelligent analysis and recognition.

[0063] The machine learning model is PSO-DMTSNet, a time-series modeling and recognition model for dynamic masked molecular fingerprint spectra. The PSO-DMTSNet model employs a dynamic mask consistency learning strategy, combining a short-term feature extraction network with a long short-term memory network (BiLSTM), and supplemented by feature attention and temporal attention as core algorithmic logics. It analyzes the feature information on time slices in detail, uses slice information to invert the overall dynamic process of evolution, and identifies and extracts multi-timescale features such as deep dynamic correlations between cross-band spectra. The model utilizes the Particle Swarm Optimization (PSO) algorithm to automatically optimize network hyperparameters in order to analyze intensity curvature abrupt change points and cross-band peak time difference configuration sensitive parameters in the three-dimensional dynamic molecular fingerprint spectrum.

[0064] Based on the three-dimensional dynamic fingerprint data matrix constructed in the previous steps This invention constructs a dynamic masked molecular fingerprint spectral temporal model (PSO-DMTSNet) to extract and analyze features from three-dimensional dynamic molecular fingerprint spectra. When combined with spectral reconstruction and backpropagation artificial neural networks, this method can accurately identify molecular bonds within different concentration ranges. The underlying logic of this algorithm is based on the synergistic use of short-term feature extraction, long-term dependency modeling, and feature attention and temporal attention mechanisms. Specifically, the algorithm first uses a one-dimensional convolutional network to analyze and extract short-term local collision event spectral feature information on a single time slice; then, the time slice information is input into a bidirectional long short-term memory network (BiLSTM) to invert and track the overall long-term physical evolution process from fragment derivation and collision to de-excitation; based on this architecture, the feature attention mechanism accurately identifies the evolutionary correlations and energy transfer paths between cross-element spectra by assigning adaptive weights to the characteristic spectral bands of different elements; while the temporal attention mechanism focuses on evaluating the importance of different evolutionary time points for macroscopic configuration identification. By comprehensively analyzing the feature information on time slices and using the slice information to invert the overall evolution process, while identifying deep correlations between spectra, this model can accurately reconstruct the mapping of "dynamic spectral sequence - high-dimensional physical evolution features - molecular space topology". Combined with a dynamic mask consistency learning strategy and automatic hyperparameter optimization using the particle swarm optimization (PSO) algorithm, this algorithm effectively solves the fundamental bottleneck of the impracticality of manually extracting dynamic correlation information in chaotic plasma radiation, filling the technical gap in macroscopic matter identification based on the evolutionary trajectory of microscopic collision networks.

[0065] For complex mixture systems, this technology achieves label-free simultaneous identification of multiple components through a time-spectral domain decoupling mechanism. Due to differences in molecular configuration, the fragmentation evolution dynamics of each component in the mixture exhibit naturally orthogonal separation in both time delay and spectral band dimensions: the characteristic radiation peaks of different components form distinguishable clusters along the time axis (e.g., the H peak of isomer A is located at...). Isomer B is located in They occupy distinct wavelength ranges on the spectral axis (such as the CN band of nitrogen-containing components and the band of pure hydrocarbon components). (Band). This orthogonality, combined with the micro-area confinement effect of the plasmonic grating, fundamentally avoids the matrix effects, plasmonic shielding effects, intermolecular coupling / polymerization interference, and inorganic background noise faced by traditional spectroscopic techniques. The AI ​​model intelligently separates the characteristics of each component in the mixed matrix by decoupling the differences in the covariance spectral distribution of the dynamic fingerprint matrix within different time windows. This configuration-modulated temporal separation mechanism, combined with the intrinsic selectivity of the spectral band, enables the simultaneous identification and quantitative inversion of multi-component substances without pre-separation.

[0066] Future technological advancements will focus on grating structure optimization and multi-scenario applications. The core upgrade direction involves introducing a third time-delay-tunable femtosecond laser. After constructing a plasma grating using a basic dual-beam system, the injection timing of the third laser beam will be precisely controlled via optical delay. This technology involves re-injecting energy during critical phases of plasma evolution. It actively modulates the branching paths of the fragment collision network, stimulating richer molecular dynamic responses, significantly improving the temporal differentiation and information entropy of the characteristic spectrum, and further enhancing its resistance to interference from complex impurity environments. Through the dynamic sensitivity of configuration encoding, this technology achieves non-destructive spatial topological identification and can be extended to multiple high-value fields. For example, in ecological environment monitoring, it enables in-situ configuration identification of pollutant isomers; in food safety supervision, it facilitates molecular topological certification of additives; in biomedical diagnostics, it precisely quantifies the spatial matching degree of DNA hydrogen bond networks; and in the field of cultural relic identification and protection, it inverts the aging trajectory of materials through molecular aging characteristics. All applications are based on the same physical core; differences in molecular spatial configuration are modulated into plasma evolution dynamics, condensing in the time-spectral domain into dynamic fingerprints that can be analyzed by intelligent algorithms.

[0067] Figure 2 This diagram illustrates the core flowchart of the femtosecond micro-area plasma molecular fingerprinting technology proposed in this application. The entire process begins with step one, which involves constructing a one-dimensional (GIBS) or two-dimensional (TIBS) plasma grating using multi-beam femtosecond laser interference. The resulting localized ultra-strong electric field is used to achieve non-thermal, stable breakdown of the sample. Step two then follows, where, within the submicron confinement channel of the grating, the broken-down molecules undergo a chain-like dynamic evolution of "dissociation-collision-excitation-emission," gradually dissociating from large molecular fragments into atomic-level fragments, radiating characteristic spectra associated with the original molecular configuration throughout this process. Step three involves simultaneously acquiring these radiation spectra and constructing a three-dimensional dynamic molecular fingerprint spectrum based on wavelength, intensity, and delay time. This allows for the complete "unfolding" of the dynamic information of molecular dissociation over time. Based on this spectrum, step four analyzes configuration-sensitive dynamic fingerprint parameters such as the dynamic linkage of specific elemental spectral lines and the timing difference of cross-elemental peaks. Finally, in step five, an AI-powered machine learning model is used to intelligently analyze and model the aforementioned dynamic fingerprints, constructing a qualitative identification network to ultimately achieve non-destructive and accurate identification of the molecular spatial topology in complex mixtures.

[0068] Compared with the prior art, the embodiments of this application have the following beneficial effects: First, it achieves high-dimensional dynamic analysis of molecular configuration, breaking through the limitations of traditional static spectroscopy. This technology captures the dynamic information of the entire process from molecular dissociation to radiation by constructing a three-dimensional dynamic molecular fingerprint spectrum of wavelength-intensity-delay time, thus avoiding the annihilation of key discrimination information by traditional time-integration spectroscopy.

[0069] Secondly, it can accurately identify spatial topological structures and effectively distinguish isomers. Since different configurations modulate fragment evolution paths and produce specific differentiations in the spectral and temporal domains, combined with machine learning to analyze dynamic correlation parameters, this method can achieve non-destructive identification of molecular spatial configurations even when the elemental composition is the same.

[0070] Third, it has the capability for in-situ analysis of complex mixtures without the need for complex sample pretreatment. The technology utilizes the micro-area confinement of the plasma grating and the temporal orthogonality of the kinetic paths of different components to directly analyze mixtures, avoiding the problems caused by matrix effects, intermolecular interference, and the need for pre-separation in traditional methods.

[0071] Fourth, it has a wide range of applications and strong scalability. Based on a unified molecular configuration dynamic coding physical kernel, this technology can be flexibly extended to multiple fields that require high-precision structural identification, such as environmental pollutant monitoring, biological tissue diagnosis, food safety, and cultural relic identification.

[0072] The following describes a possible implementation of a femtosecond micro-plasma molecular fingerprinting system, which performs the various execution steps and corresponding technical effects of the femtosecond micro-plasma molecular fingerprinting method shown in the above embodiments and possible implementations, including: The femtosecond laser generator module is used to generate multiple femtosecond laser beams and control their interference to form a plasma grating; The sample micro-area interaction module is used to hold the sample to be tested and provide a micro-area breakdown environment under the constraint of a plasma grating. The time-domain resolved spectral acquisition module, including an ICCD detector and a spectrometer, is used to synchronously acquire emission spectra at different delay times; The data processing and recognition module has a built-in memory and processor. The processor is configured to execute instructions to construct a three-dimensional dynamic molecular fingerprint spectrum, extract dynamic correlation parameters, and run a machine learning model to output molecular spatial topological configuration recognition results.

[0073] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A femtosecond micro-region plasma molecular fingerprinting method, characterized in that, include: Multiple femtosecond laser beams are focused and interfered in a micro-region of the sample to form a plasma grating. A localized ultra-strong electric field is generated in the antinode region of the plasma grating, thereby breaking down the chemical bonds in the sample in a non-thermal manner. By utilizing the spatial confinement effect of a plasma grating, a high-density plasma channel is formed and maintained in the antinode region. The generated debris undergoes a chain evolution process within the channel and emits characteristic spectra associated with the molecular spatial configuration. Simultaneously collect emission spectra emanating from the plasma channel at different delay times to construct a three-dimensional dynamic molecular fingerprint spectrum encoding fragment evolution dynamics based on wavelength, intensity, and delay time. From the three-dimensional dynamic molecular fingerprint spectrum, dynamic correlation parameters of multiple comprehensive evolution laws, including intensity envelope evolution, peak time difference of cross-element spectral peaks and linewidth evolution, are extracted. Machine learning models are used to analyze dynamic correlation parameters in order to reconstruct and identify the spatial topological configuration of molecules in the sample to be tested.

2. The femtosecond micro-area plasma molecular fingerprinting method as described in claim 1, characterized in that, The steps of using multiple femtosecond laser beams to focus and interfere in a micro-region of the sample to form a plasma grating include: A femtosecond laser system with a center wavelength of 800 nm, a pulse width of 80 fs, and a single pulse energy of 2 mJ was used. Based on the femtosecond laser system, multiple femtosecond laser beams are focused onto the surface of the sample to be tested, and a self-focusing effect occurs in the air medium. When the light intensity reaches the multiphoton ionization threshold, plasma is formed. A dynamic balance is achieved between the self-focusing effect and the plasma defocusing effect to form a plasma single-light filament; Split at least two femtosecond laser beams at a spatial angle Incident and interference occur, forming a period of The plasma grating, in which , The wavelength of the laser; The plasma grating, through coherent superposition, generates a periodic electric field enhancement in the antinode region, thereby increasing the local light intensity to [a certain level]. Quantity, of which The number of coherent beams. This represents the saturation intensity of a single filament.

3. The femtosecond micro-area plasma molecular fingerprinting method as described in claim 1, characterized in that, When two femtosecond laser beams are intersected to generate periodic gain fringes, a two-dimensional transient plasma grating structure is formed, which is labeled as a plasma grating-induced breakdown spectrum (GIBS) structure. When nonlinear interference is generated using three high-intensity femtosecond laser filaments, a transient plasma grating channel is established, which is labeled as the TIBS structure of the three-filament coupled transient interference spectroscopy method.

4. The femtosecond micro-area plasma molecular fingerprinting method as described in claim 1, characterized in that, The generated fragments undergo a chain-like evolutionary process within the channel, including: Under the influence of a localized ultra-strong electric field, molecular bonding electrons are instantaneously extracted through the quantum tunneling effect, triggering avalanche ionization and forming a high-density electron cluster. Extreme charge imbalance triggers a violent Coulomb explosion, which explosively dissociates the molecular structure into atomic ions and molecular fragments. The generation path of these fragments is related to the spatial topology and bond energy distribution of the original molecule. The dissociated fragments, confined in the submicron space of the plasma channel, undergo continuous inelastic collisions with high-density electrons and are excited, subsequently emitting characteristic spectra through radiative transitions.

5. The femtosecond micro-area plasma molecular fingerprinting method as described in claim 1, characterized in that, The steps involved in constructing a wavelength-intensity-delay time three-dimensional dynamic molecular fingerprint spectrum encoding fragment evolution dynamics include: An enhanced charge-coupled device (ICCD) detector with a time-domain gate width ≤ 1 ns was used to acquire spectral signals; Construct a three-dimensional matrix of wavelength-intensity-delay time As a three-dimensional dynamic molecular fingerprint spectrum, in which To delay time, For wavelength, The light intensity is represented by the three-dimensional dynamic molecular fingerprint spectrum, which carries information on the energy levels of fragment electronic state transitions in the wavelength dimension, records the distribution of non-equilibrium particle numbers and fragment abundance in the intensity dimension, and encodes the reaction path and de-excitation evolution phase in the delay time dimension.

6. The femtosecond micro-area plasma molecular fingerprinting method as described in claim 1, characterized in that, The characteristic spectra include the hydrogen atom line H 656.3 nm, the carbon atom line C 247.8 nm, the oxygen atom line O 777.2 nm, the nitrogen atom line N 746.8 nm, and the cyano radical band CN 388.3 nm.

7. The femtosecond micro-area plasma molecular fingerprinting method as described in claim 1, characterized in that, Dynamic correlation parameters are used to characterize the triple coupling mapping relationship between molecular configuration and spectral evolution, including: The dynamic relationship between the evolution of elemental spectral line intensity and linewidth; The set of time differences between the peak occurrence times of characteristic spectral peaks of different elements; Global characteristics of the high-dimensional correlation matrix of the full-band spectrum co-evolving over time.

8. The femtosecond micro-area plasma molecular fingerprinting method as described in claim 1, characterized in that, The machine learning model is PSO-DMTSNet, a time-series modeling and recognition model for dynamic masked molecular fingerprint spectra. The PSO-DMTSNet model employs a dynamic mask consistency learning strategy, combining a short-term feature extraction network with a long short-term memory network (BiLSTM), and supplemented by feature attention and temporal attention as core algorithmic logics. It analyzes the feature information on time slices in detail, uses slice information to inverse the overall dynamic process of evolution, and identifies and extracts features at multiple time scales. The PSO-DMTSNet model utilizes the Particle Swarm Optimization (PSO) algorithm to automatically optimize network hyperparameters in order to resolve intensity curvature abrupt change points and cross-band peak time difference configuration sensitive parameters in three-dimensional dynamic molecular fingerprint spectra.

9. A femtosecond micro-area plasma molecular fingerprinting system, used to perform the method as described in any one of claims 1-8, characterized in that, include: The femtosecond laser generator module is used to generate multiple femtosecond laser beams and control their interference to form a plasma grating; The sample micro-area interaction module is used to hold the sample to be tested and provide a micro-area breakdown environment under the constraint of a plasma grating. The time-domain resolved spectral acquisition module, including an ICCD detector and a spectrometer, is used to synchronously acquire emission spectra at different delay times; The data processing and recognition module has a built-in memory and processor. The processor is configured to execute instructions to construct a three-dimensional dynamic molecular fingerprint spectrum, extract dynamic correlation parameters, and run a machine learning model to output molecular spatial topological configuration recognition results.

Images