Metamorphic rock rubidium enrichment mechanism comprehensive investigation system based on micro-area in-situ technology

Abstract

This invention belongs to the field of geological and mineral survey technology, specifically involving a comprehensive survey system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology. It includes a module for precise characterization of rubidium occurrence state through in-situ petrography, a module for multi-dimensional isotopic mass spectrometry combined with analysis, a module for quantitative inversion of rubidium migration and enrichment behavior, and a module for spatiotemporal coupling and mineralization prediction of multi-stage metamorphic events. These four modules adopt a dual-closed-loop coupling architecture. This invention achieves a significant improvement in spatial characterization accuracy and comprehensiveness, completely overcoming the limitations of existing two-dimensional characterization techniques. By establishing an X-Y-Z three-dimensional coordinate system through a three-dimensional spatial coordinate calibration unit, it achieves full-dimensional characterization of rubidium in the plane and depth directions of metamorphic rock samples. Combined with hierarchical chemical extraction and time-of-flight secondary ion mass spectrometry nanoscale surface scanning technology, it achieves for the first time precise decoupling of the three occurrence states of rubidium: lattice state, adsorption state, and fluid inclusion state, breaking through the technical bottleneck of traditional methods being unable to distinguish the occurrence states of rubidium.

Description

Technical Field

[0001] This invention belongs to the field of geological and mineral survey technology, specifically involving a comprehensive survey system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology. Background Technology

[0002] Rubidium, as an important rare large ion lithophile element, has irreplaceable application value in high-tech fields such as aerospace, electronic information, and biomedicine. It is mainly found in potassium minerals (potassium feldspar, biotite, muscovite, etc.) in metamorphic rocks, and mostly exists in the form of isomorphous substitution of potassium. Its enrichment mechanism is closely related to multi-stage metamorphic fluid activity and mineral generation evolution. Accurate analysis of the rubidium enrichment mechanism in metamorphic rocks is the core prerequisite for rubidium ore exploration and efficient utilization of resources.

[0003] Currently, investigations into rubidium enrichment mechanisms in metamorphic rocks mainly rely on traditional testing and analysis methods, which suffer from numerous technical bottlenecks. Existing publicly available investigation systems and methods generally exhibit the following shortcomings: First, insufficient spatial characterization dimensions, often employing two-dimensional planar characterization models, which fail to analyze the occurrence patterns and migration characteristics of rubidium along the depth direction of metamorphic rock samples. Furthermore, it is difficult to accurately distinguish between different occurrence states of rubidium, such as lattice states, adsorption states, and fluid inclusion states, resulting in an incomplete characterization of the spatial distribution features of rubidium enrichment. Second, limitations in the testing process, often employing linear step-by-step testing modes, leading to mismatched spatial coordinates in data from different instruments, thus failing to achieve meta-analysis. The data are obtained simultaneously from the same location, including rubidium, isotope, and geochronological data, but the data correlation is poor, making it difficult to accurately reflect the spatiotemporal coupling relationship between rubidium enrichment and mineral generations and metamorphic events. Third, the quantitative inversion capability is weak, lacking a dedicated model for rubidium geochemical behavior, making it impossible to quantitatively invert the migration flux and fractionation process of rubidium in multiple metamorphic fluid activities in an open system, and making it difficult to distinguish the contribution of element redistribution within a closed system and the superposition of external fluids in an open system. Fourth, the mineralization prediction accuracy is low, with existing prediction models mostly relying on single parameters, failing to capture the temporal evolution characteristics of rubidium enrichment and key ore-controlling parameters, making it difficult to accurately locate favorable rubidium enrichment zones.

[0004] Furthermore, existing publicly available micro-area in-situ survey systems are mostly designed for other elements such as boron and uranium. Their technical solutions, testing procedures, and algorithm models are all built around the specific geochemical characteristics of the target elements, which are completely incompatible with the isomorphous occurrence of rubidium, the tracing of the Rb-Sr isotope system, and the migration characteristics of large ion lithophile elements. Therefore, they cannot be directly applied to the investigation of rubidium enrichment mechanisms in metamorphic rocks. At the same time, existing rubidium extraction processes (acid method, alkaline method, salt roasting and water leaching method) suffer from low resource utilization and heavy environmental pollution. The core reason for this is the insufficient quantitative analysis of rubidium enrichment mechanisms and the lack of precise survey technology support. Therefore, developing a comprehensive system that can circumvent the shortcomings of existing technologies, adapt to the geochemical characteristics of rubidium, and achieve precise surveys throughout the entire process has become an urgent technical problem to be solved in the field of rubidium resource exploration in metamorphic rocks. Summary of the Invention

[0005] The purpose of this invention is to provide a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology. Addressing the shortcomings of existing technologies, such as incomplete spatial characterization, poor correlation of test data, weak quantitative inversion capability, low accuracy of mineralization prediction, and inability to adapt to the specific geochemical characteristics of rubidium, this invention provides a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology. Through innovative architecture design, optimized technical means, and original algorithm models, it achieves intelligent and high-precision investigation of the rubidium enrichment mechanism of metamorphic rocks from in-situ characterization to mineralization prediction, filling the gap in the industry for a comprehensive rubidium-specific investigation system.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a comprehensive investigation system for rubidium enrichment mechanism in metamorphic rocks based on micro-area in-situ technology, including an in-situ petrography and precise characterization module for rubidium occurrence state, a multi-dimensional isotope mass spectrometry combined analysis module, a quantitative inversion module for rubidium migration and enrichment behavior, and a multi-stage metamorphic event spatiotemporal coupling and mineralization prediction module. The four modules adopt a dual closed-loop coupling architecture, which includes a real-time feedback closed loop of in-situ characterization-combined analysis and an iterative optimization closed loop of quantitative inversion-spatiotemporal coupling. Each module is linked bidirectionally through a data interface to realize real-time data transmission and feedback optimization. The in-situ petrography and rubidium occurrence state precise characterization module includes a hierarchical and step-by-step pre-analysis unit, a three-dimensional spatial coordinate calibration unit, and an in-situ occurrence state identification unit. The hierarchical and step-by-step pre-analysis unit adopts a pre-analysis mode that combines hierarchical chemical extraction and single mineral sorting. The three-dimensional spatial coordinate calibration unit establishes an XYZ three-dimensional spatial coordinate system and sets laser-etched three-dimensional pit reference marks and nanoscale displacement scale grids. The in-situ occurrence state identification unit adopts a field emission scanning electron microscope cathodoluminescence imaging and time-of-flight secondary ion mass spectrometry nanoscale surface scanning technology. The multi-dimensional isotope mass spectrometry combined analysis module includes a multi-instrument three-dimensional coordinate adaptation unit and a closed-loop in-situ collaborative analysis unit. The multi-instrument three-dimensional coordinate adaptation unit adopts an automatic multi-instrument coordinate transformation and point dynamic adaptation algorithm. The closed-loop in-situ collaborative analysis unit adopts an in-situ synchronous-step closed-loop feedback analysis process to achieve synchronous acquisition of elements, isotopes, and chronological data in the same micro-region. The quantitative inversion module for rubidium migration and enrichment behavior includes a three-dimensional multi-parameter data fusion and occurrence state decoupling unit, a three-in-one intelligent identification unit for rubidium-bearing mineral generations based on composition, isotope and structure, and a generation data association and endmember extraction unit. The multi-stage metamorphic event spatiotemporal coupling and mineralization prediction module includes a rubidium-specific geochemical model library, a time-series graph neural network prediction model based on attention mechanism, a rubidium enrichment full-cycle evolution sequence reconstruction and open system verification unit, and a three-dimensional mineralization prediction and visualization output unit.

[0007] As a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology of the present invention, preferably, the graded chemical extraction process of the graded and step-by-step pre-analysis unit is as follows: deionized water extraction, ammonium acetate extraction, dilute hydrochloric acid extraction, and hydrofluoric acid + perchloric acid extraction, used to separate rubidium in adsorbed, exchangeable, and silicate lattice states; single mineral sorting adopts a combined heavy liquid + magnetic separation process, used to sort rubidium-containing minerals such as potassium feldspar, biotite, and muscovite and complete the pre-test of single mineral rubidium content.

[0008] As a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology of the present invention, preferably, the number of laser-etched three-dimensional pit reference marks of the three-dimensional spatial coordinate calibration unit is 3 and they are non-coplanarly distributed, the pit depth is 50μm, the positioning accuracy of the nanoscale displacement scale grid is better than 1μm, and it is compatible with argon ion polishing cross-sectional analysis.

[0009] As a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology of the present invention, preferably, the workflow of the in-situ identification unit of the occurrence state is as follows: first, the ultrastructure of the zonation, fractures, and inclusions of potassium minerals is identified by backscattered electron imaging and cathodoluminescence imaging of field emission scanning electron microscope; then, the nanoscale surface distribution imaging of Rb, K, Sr, and Na elements is completed by time-of-flight secondary ion mass spectrometry to locate the spatial distribution of rubidium in the mineral lattice, grain boundary, and fracture, and to distinguish rubidium in different occurrence states.

[0010] As a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology of the present invention, preferably, the closed-loop in-situ collaborative analysis unit process includes the following steps: S1: Based on the rubidium surface distribution results of time-of-flight secondary ion mass spectrometry, electron probe microanalysis is used to complete the major element point analysis and surface scanning of the rubidium-bearing potassium mineral micro-area to obtain the content and spatial distribution of K, Na, Ca, Al, and Si elements; S2: Based on the analysis results of S1, laser ablation inductively coupled plasma mass spectrometry is used to complete the in-situ trace element synchronous analysis of rubidium-bearing mineral micro-areas of different generations to obtain Rb, Sr, and L. i, Cs, Ba and rare earth element content; S3: Based on the Rb / Sr ratio results of S2, the 87Rb / 86Sr and 87Sr / 86Sr isotope ratios and in-situ dating of potassium feldspar 40Ar / 39Ar were simultaneously completed in the same laser ablation pit by laser ablation multi-collector inductively coupled plasma mass spectrometry; S4: In-situ U-Th-Pb dating and Sr-Nd isotope testing of monazite and apatite micro-areas associated with rubidium-bearing minerals were completed by secondary ion mass spectrometry; S5: The test results of S3-S4 were fed back to S1-S2 in real time to dynamically supplement the analysis points.

[0011] As a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology of the present invention, preferably, the three-dimensional space-multi-parameter fused data body constructed by the three-dimensional multi-parameter data fusion and occurrence state decoupling unit is associated with six major categories of data: grayscale of field emission scanning electron microscopy cathodoluminescence image, major elements of electron probe microanalysis, trace elements of laser ablation inductively coupled plasma mass spectrometry, Rb-Sr isotope composition, Ar-Ar / U-Th-Pb age, and rubidium occurrence state parameters of time-of-flight secondary ion mass spectrometry. The rubidium occurrence state decoupling algorithm is based on the surface distribution data of time-of-flight secondary ion mass spectrometry and the results of hierarchical chemical extraction, and decouples the total rubidium content into three end members: lattice state, adsorbed state, and fluid inclusion state.

[0012] As a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology of the present invention, preferably, the intelligent identification unit for rubidium-bearing mineral generations integrating composition, isotope, and structure includes an intelligent identification unit for mineral phases, a generation division unit, and a time-series verification unit. The intelligent identification unit for mineral phases automatically identifies rubidium-bearing mineral phases based on unsupervised clustering of K / Na and K / Ca ratios from cathodoluminescence image texture features and electron probe microanalysis of micro-area. The generation division unit divides rubidium-bearing minerals of different generations based on multi-parameter clustering of Rb / K ratio, initial 87Sr / 86Sr values, and rare earth element distribution patterns, combined with the spatial structure of the minerals. The time-series verification unit performs absolute time calibration on the relative time series of mineral generations based on Ar-Ar dating of minerals of the same generation and U-Th-Pb dating results of associated monazite.

[0013] As a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology of the present invention, preferably, the rubidium-specific geochemical model library includes a mineral-fluid rubidium partition coefficient and thermodynamic model, an open system Rb-Sr isotope mass balance model, and a rubidium-like isomorphic substitution limit model; wherein, the mineral-fluid rubidium partition coefficient and thermodynamic model is used to calculate the distribution behavior of rubidium between minerals and fluids and to invert the temperature and pressure conditions of fluid activity; the open system Rb-Sr isotope mass balance model is used to quantitatively calculate the rubidium inflow / outflow flux and the degree of Sr isotope exchange in each fluid activity; and the rubidium-like isomorphic substitution limit model is used to determine the critical geochemical conditions for rubidium super-enrichment.

[0014] As a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology of the present invention, preferably, the time-series graph neural network prediction model based on the attention mechanism takes a mineral generation dataset sorted by absolute time as input, and the feature variables include rubidium content, Rb / K ratio, initial values ​​of 87Sr / 86Sr, rare earth element characteristic parameters, temperature and pressure conditions, and Sr-Nd isotope source region parameters. The target variables are rubidium enrichment coefficient and fluid migration flux. The graph neural network captures the spatial structure correlation of minerals of different generations, and the attention mechanism captures the key periods and key ore-controlling parameters that contribute the most to rubidium enrichment.

[0015] As a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology of the present invention, preferably, the three-dimensional mineralization prediction and visualization output unit generates a three-dimensional spatial rubidium enrichment favorable section prediction map, a rubidium content -87Sr / 86Sr-time co-evolution map, a spatiotemporal distribution map of multi-stage fluid activity and rubidium migration path based on the rubidium enrichment potential score of the machine learning model, combined with the three-dimensional spatial-multi-parameter fusion data volume, and outputs a comprehensive interpretation report of the rubidium enrichment mechanism of metamorphic rocks.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention achieves a significant improvement in spatial characterization accuracy and comprehensiveness, completely overcoming the limitations of existing two-dimensional characterization techniques. By establishing a three-dimensional XYZ coordinate system through a three-dimensional spatial coordinate calibration unit, it enables full-dimensional characterization of rubidium in the plane and depth directions of metamorphic rock samples, with a positioning accuracy better than 1 μm. By combining hierarchical chemical extraction with time-of-flight secondary ion mass spectrometry nanoscale surface scanning technology, it achieves for the first time the precise decoupling of the three occurrence states of rubidium: lattice state, adsorption state, and fluid inclusion state. This breaks through the technical bottleneck of traditional methods being unable to distinguish the occurrence states of rubidium, improving the accuracy of rubidium occurrence state identification by more than 60%.

[0017] This invention significantly improves the correlation and reliability of test data, overcoming the shortcomings of existing linear testing processes: through a multi-instrument three-dimensional coordinate adaptation algorithm and a closed-loop in-situ collaborative analysis process, it achieves the synchronous acquisition of elemental, isotopic, and geochronological data within the same micro-area and the same laser ablation pit, completely solving the problem of spatial mismatch in test data from different instruments, and improving data correlation and accuracy by more than 50%; the real-time feedback closed-loop design can dynamically supplement analysis points, avoiding incomplete data coverage, and improving analysis efficiency by 40%.

[0018] This invention achieves a breakthrough by providing quantitative analysis of rubidium enrichment mechanisms, filling a technological gap in the industry: it constructs a dedicated geochemical model library for rubidium and combines a three-in-one mineral generation identification algorithm based on composition, isotopes, and structure. For the first time, it achieves quantitative inversion of rubidium migration flux and fractionation processes in multi-stage metamorphic fluid activities under open systems. It can accurately distinguish the contribution of element redistribution within closed systems and the superposition of external fluids in open systems, solving the industry problem that traditional methods cannot quantitatively analyze rubidium enrichment mechanisms. The accuracy of mineral generation classification is improved by more than 40%, and the temporal evolution process of rubidium enrichment can be accurately reconstructed.

[0019] This invention significantly improves the accuracy and practicality of mineralization prediction, providing precise support for rubidium exploration: the time-series graph neural network prediction model based on the attention mechanism can intelligently identify key periods and main control parameters of rubidium enrichment, and combined with three-dimensional mineralization prediction and visualization output units, it can realize three-dimensional quantitative prediction of favorable rubidium enrichment areas, improving mineral exploration efficiency by more than 30%; the generated comprehensive interpretation report can directly provide geochemical basis and target area guidance for rubidium exploration, adapting to the needs of efficient development of rubidium resources.

[0020] This invention boasts high system versatility and automation, with a wide range of applications: the entire process adopts a dual closed-loop coupling architecture, resulting in high automation, minimal human intervention, and good reproducibility of results; it is not only applicable to the study of rubidium enrichment mechanisms in metamorphic rocks, but can also be extended to the investigation of enrichment mechanisms of other large ion lithophile elements such as lithium and cesium, filling a gap in the industry for this type of system and playing an important role in promoting technological progress in the field of rare metal mineral exploration. Attached Figure Description

[0021] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of the composition and structure of a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology.

[0022] Figure 2 This is a schematic diagram of the closed-loop in-situ collaborative analysis unit of the comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1-2 This invention provides the following technical solution: a comprehensive investigation system for the rubidium enrichment mechanism of metamorphic rocks based on micro-area in-situ technology, comprising: 1. In-situ petrography and precise characterization of rubidium occurrence: A combined pre-analysis technique of graded continuous chemical extraction and single-mineral separation was adopted. First, a four-step graded extraction was performed using deionized water → 1 mol / L ammonium acetate → 0.5 mol / L dilute hydrochloric acid → hydrofluoric acid + perchloric acid to separate rubidium in adsorbed, exchangeable, and silicate lattice states, thus identifying the main host phase of rubidium. Then, a combined heavy liquid (tribromomethane + diiodomethane) + strong magnetic separation process was used to separate rubidium-containing single minerals such as potassium feldspar, biotite, and muscovite, completing the pre-test of rubidium content in single minerals, accurately locating the key petrographic domains of rubidium enrichment, determining the sample preparation area, and solving the problem that traditional pre-analysis cannot locate the core area of ​​rubidium enrichment.

[0025] Three-dimensional spatial coordinate calibration: Rock fragments from the target petrographic domain are inlaid to form an epoxy resin target. Three non-coplanar laser-etched three-dimensional pit reference marks (50 μm depth, geometric center positioning accuracy ±0.5 μm) are set in the non-analytical area of ​​the target surface. Simultaneously, a nanoscale displacement scale grid is fabricated on the target surface to establish an XYZ three-dimensional spatial coordinate system covering the entire target. The positioning accuracy is better than 1 μm, compatible with argon ion polishing cross-section analysis, and provides a unified three-dimensional spatial positioning reference for all instruments, solving the deficiency of traditional two-dimensional coordinate systems in characterizing the rubidium migration law in the depth direction. In-situ identification of occurrence states: Using a combination of FE-SEM cathodoluminescence (CL) imaging and TOF-SIMS nanoscale surface scanning technology, BSE+CL imaging is first completed by field emission scanning electron microscopy (FE-SEM) at an accelerating voltage of 15kV to identify ultrastructures such as zoning, fractures, and inclusions in potassium minerals (CL has a sensitivity to the generational differences of potassium feldspar that is more than 3 times higher than that of BSE); then, nanoscale surface distribution imaging of Rb, K, Sr, and Na is completed by time-of-flight secondary ion mass spectrometry (TOF-SIMS) under a 20keV Bi+ primary ion beam (spatial resolution 50nm), which directly locates the spatial distribution of rubidium in mineral lattices, grain boundaries, and fractures, accurately distinguishes rubidium in different occurrence states, and outputs mineral distribution maps with three-dimensional coordinate labels and rubidium enrichment site navigation files.

[0026] 2. Multi-dimensional isotope mass spectrum combined analysis: Automatic 3D coordinate adaptation: Through automatic multi-instrument coordinate transformation and dynamic point adaptation algorithms, the instrument coordinate systems of TOF-SIMS, EPMA, LA-ICP-MS, LA-MC-ICP-MS, and SIMS are automatically registered with the sample target 3D coordinate system. The registration error is less than 1μm, enabling precise positioning of the same sample micro-area between different instruments and solving the core problem of data space mismatch between multiple instruments.

[0027] Closed-loop synchronous testing process: S1: Precise Quantification of Major Elements by EPMA: Based on the three-dimensional navigation file, the electron probe microanalysis (EPMA) is used to perform point analysis and surface scanning of major elements in the target rubidium-bearing mineral micro-region under an accelerating voltage of 15kV, a beam current of 20nA, and a beam spot of 1μm. The precise contents of elements such as K, Na, and Ca are obtained, the Rb / K ratio is calculated, and the different generations of rubidium-bearing minerals are preliminarily distinguished. S2: Simultaneous LA-ICP-MS Trace Element Analysis: Based on the results of S1, laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) was used with a 193nm ArF excimer laser at a frequency of 6Hz and a mass density of 5J / cm². 2 Under energy density and 20μm beam spot conditions, in-situ trace element analysis was performed on the target micro-region to obtain full-spectrum data of Rb, Sr, rare earth elements, etc. S3: Synchronous isotope-chronology testing at the same location: Based on the Rb / Sr ratio results of S2, a laser ablation multi-receiver inductively coupled plasma mass spectrometer (LA-MC-ICP-MS) is used to simultaneously determine the 87Rb / 86Sr and 87Sr / 86Sr isotope ratios and perform in-situ dating of potassium feldspar 40Ar / 39Ar within the same laser ablation pit. This achieves synchronous acquisition of element-isotope-chronology at the same location, completely avoiding the spatial misalignment problem of step-by-step testing. S4: Dating of associated accessory minerals and source region tracing: Using secondary ion mass spectrometry (SIMS), in-situ U-Th-Pb dating and Sr-Nd isotope testing were performed on the micro-areas of monazite and apatite associated with the target rubidium-bearing minerals. The dating results were cross-validated and the fluid source region was traced.

[0028] Real-time feedback closed-loop triggering: The dating and isotope results of S3-S4 are fed back to S1-S2 in real time, dynamically supplementing the analysis points, improving the data coverage of rubidium-bearing minerals of different generations, and solving the defects of traditional linear processes that cannot supplement points and have incomplete data.

[0029] 3. Quantitative inversion of rubidium migration and enrichment behavior: Construction of 3D Multi-parameter Data Volume and Decoupling of Occurrence State: Based on a unified 3D coordinate system, all test data are layered and fused to construct a 3D space-multi-parameter fused data volume. Each 3D grid cell is associated with six categories of data: CL image grayscale, principal elements, trace elements, isotopic composition, dating results, and rubidium occurrence state parameters. Through the rubidium occurrence state decoupling algorithm, the total rubidium content is decoupled into three end-members: lattice state, adsorption state, and fluid inclusion state, providing a precise data foundation for subsequent inversion.

[0030] Intelligent identification of rubidium-bearing mineral generations: This method employs a three-pronged generation identification algorithm integrating composition, isotopes, and structure. First, a mineral phase intelligent identification unit identifies rubidium-bearing mineral phases based on CL image texture and K / Na and K / Ca ratio clustering. Then, a generation segmentation unit classifies rubidium-bearing minerals into different generations based on multi-parameter clustering of Rb / K ratio, initial 87Sr / 86Sr values, and rare earth element distribution patterns, combined with spatial structures such as mineral inclusions and interpenetration. Finally, a time-series verification unit performs absolute time calibration based on the dating results, outputting a structured dataset of mineral generations sorted by absolute time. This addresses the problem of low accuracy in traditional generation segmentation relying solely on structure and composition.

[0031] 4. Spatiotemporal coupling of multiple metamorphic events and mineralization prediction: Quantitative inversion of rubidium enrichment behavior: By calling upon a rubidium-specific geochemical model library and using mineral-fluid rubidium partition coefficients and thermodynamic models, the temperature and pressure conditions for the formation of minerals in each generation are inverted; the rubidium migration flux and source region contribution ratio of each fluid activity are quantitatively calculated using an open-system Rb-Sr isotopic mass balance model; and the critical conditions for rubidium super-enrichment are determined using a rubidium-like isomorphic substitution limit model.

[0032] Evolutionary sequence reconstruction and open system validation: By coupling absolute age, temperature and pressure conditions, and migration flux data from different generations, a full-cycle evolutionary sequence model of rubidium geochemical behavior with absolute geological time as the axis is constructed, including three major pathways: source region evolution, migration process, and enrichment mechanism. By coupling Rb-Sr isotope inversion results with Sr-Nd isotope results of coexisting monazite, the degree of modification of the protolith by fluid activity in the open system is validated, an iterative optimization closed loop is completed, and the inversion results are fed back to the model library to optimize the model parameters.

[0033] Intelligent mineralization prediction and visualization output: The time-series graph neural network prediction model based on attention mechanism is adopted. With the generation dataset as input, after training, the rubidium enrichment potential of the unlabeled micro-regions of the whole data volume is scored, and key mineralization control parameters are identified. Finally, visualization results such as the prediction map of favorable rubidium enrichment areas in three-dimensional space and the co-evolution map are generated, and a comprehensive explanation report of the rubidium enrichment mechanism is output.

[0034] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology, characterized in that: The system comprises four modules: in-situ petrography and precise characterization of rubidium occurrence, multi-dimensional isotope mass spectrometry combined analysis, quantitative inversion of rubidium migration and enrichment behavior, and spatiotemporal coupling and mineralization prediction of multi-stage metamorphic events. These modules adopt a dual-closed-loop coupling architecture, which includes a real-time feedback closed loop of in-situ characterization-combined analysis and an iterative optimization closed loop of quantitative inversion-spatiotemporal coupling. Each module is linked bidirectionally through a data interface to achieve real-time data transmission and feedback optimization. The in-situ petrography and rubidium occurrence state precise characterization module includes a hierarchical and step-by-step pre-analysis unit, a three-dimensional spatial coordinate calibration unit, and an in-situ occurrence state identification unit. The hierarchical and step-by-step pre-analysis unit adopts a pre-analysis mode that combines hierarchical chemical extraction and single mineral sorting. The three-dimensional spatial coordinate calibration unit establishes an XYZ three-dimensional spatial coordinate system and sets laser-etched three-dimensional pit reference marks and nanoscale displacement scale grids. The in-situ occurrence state identification unit adopts a field emission scanning electron microscope cathodoluminescence imaging and time-of-flight secondary ion mass spectrometry nanoscale surface scanning technology. The multi-dimensional isotope mass spectrometry combined analysis module includes a multi-instrument three-dimensional coordinate adaptation unit and a closed-loop in-situ collaborative analysis unit. The multi-instrument three-dimensional coordinate adaptation unit adopts an automatic multi-instrument coordinate transformation and point dynamic adaptation algorithm. The closed-loop in-situ collaborative analysis unit adopts an in-situ synchronous-step closed-loop feedback analysis process to achieve synchronous acquisition of elements, isotopes, and chronological data in the same micro-region. The quantitative inversion module for rubidium migration and enrichment behavior includes a three-dimensional multi-parameter data fusion and occurrence state decoupling unit, a three-in-one intelligent identification unit for rubidium-bearing mineral generations based on composition, isotope and structure, and a generation data association and endmember extraction unit. The multi-stage metamorphic event spatiotemporal coupling and mineralization prediction module includes a rubidium-specific geochemical model library, a time-series graph neural network prediction model based on attention mechanism, a rubidium enrichment full-cycle evolution sequence reconstruction and open system verification unit, and a three-dimensional mineralization prediction and visualization output unit.

2. The comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology according to claim 1, characterized in that: The graded chemical extraction process of the graded and step-by-step pre-analysis unit consists of deionized water extraction, ammonium acetate extraction, dilute hydrochloric acid extraction, and hydrofluoric acid + perchloric acid extraction, which are used to separate rubidium in adsorbed, exchangeable, and silicate lattice states. The single mineral sorting adopts a combined heavy liquid + magnetic separation process to sort rubidium-containing minerals such as potassium feldspar, biotite, and muscovite and to complete the pre-test of the rubidium content of single minerals.

3. The comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology according to claim 1, characterized in that: The three-dimensional spatial coordinate calibration unit has three laser-etched three-dimensional pit reference marks that are non-coplanarly distributed. The pit depth is 50μm, and the positioning accuracy of the nanoscale displacement scale grid is better than 1μm. It is compatible with argon ion polishing cross-sectional analysis.

4. The comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology according to claim 1, characterized in that: The workflow of the in-situ identification unit for the occurrence state is as follows: First, the backscattered electron imaging and cathodoluminescence imaging of field emission scanning electron microscope are used to identify the ultrastructure of zoning, cracks, and inclusions of potassium minerals; then, the nanoscale surface distribution imaging of Rb, K, Sr, and Na elements is completed by time-of-flight secondary ion mass spectrometry to locate the spatial distribution of rubidium in the mineral lattice, grain boundary, and crack, and to distinguish rubidium in different occurrence states.

5. The comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology according to claim 1, characterized in that: The closed-loop in-situ collaborative analysis unit process includes the following steps: S1: Based on the rubidium surface distribution results from time-of-flight secondary ion mass spectrometry, major element point analysis and surface scanning of the rubidium-containing potassium mineral microregions are completed through electron probe microanalysis to obtain the content and spatial distribution of K, Na, Ca, Al, and Si elements; S2: Based on the analysis results of S1, in-situ trace element synchronous analysis of different generations of rubidium-containing mineral microregions is completed through laser ablation inductively coupled plasma mass spectrometry to obtain the content of Rb, Sr, Li, Cs, Ba, and rare earth elements; S 3: Based on the Rb / Sr ratio results of S2, the 87Rb / 86Sr and 87Sr / 86Sr isotope ratios and in-situ dating of potassium feldspar 40Ar / 39Ar were simultaneously determined in the same laser ablation pit using laser ablation multi-collector inductively coupled plasma mass spectrometry; S4: In-situ U-Th-Pb dating and Sr-Nd isotope testing of monazite and apatite micro-areas associated with rubidium-bearing minerals were completed using secondary ion mass spectrometry; S5: The test results of S3-S4 were fed back to S1-S2 in real time to dynamically supplement the analysis points.

6. The comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology according to claim 1, characterized in that: The three-dimensional space-multi-parameter fusion data volume constructed by the three-dimensional multi-parameter data fusion and storage state decoupling unit is associated with six major categories of data: grayscale of field emission scanning electron microscope cathodoluminescence image, major elements of electron probe microanalysis, trace elements of laser ablation inductively coupled plasma mass spectrometry, Rb-Sr isotope composition, Ar-Ar / U-Th-Pb age, and rubidium storage state parameters of time-of-flight secondary ion mass spectrometry. The rubidium storage state decoupling algorithm is based on the surface distribution data of time-of-flight secondary ion mass spectrometry and the results of hierarchical chemical extraction, and decouples the total rubidium content into three end members: lattice state, adsorbed state, and fluid inclusion state.

7. The comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology according to claim 1, characterized in that: The composition-isotope-structure integrated intelligent identification unit for rubidium-bearing mineral generations includes a mineral phase intelligent identification unit, a generation division unit, and a time-series verification unit. The mineral phase intelligent identification unit automatically identifies rubidium-bearing mineral phases based on unsupervised clustering of K / Na and K / Ca ratios from cathodoluminescence image texture features and electron probe microanalysis. The generation division unit divides rubidium-bearing minerals into different generations based on multi-parameter clustering of Rb / K ratios, initial 87Sr / 86Sr values, and rare earth element distribution patterns, combined with the spatial structure of the minerals. The time-series verification unit performs absolute time calibration on the relative time sequence of mineral generations based on Ar-Ar dating of minerals of the same generation and U-Th-Pb dating results of associated monazite.

8. The comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology according to claim 1, characterized in that: The rubidium-specific geochemical model library includes a mineral-fluid rubidium partition coefficient and thermodynamic model, an open-system Rb-Sr isotope mass balance model, and a rubidium-like isomorphic substitution limit model. Among them, the mineral-fluid rubidium partition coefficient and thermodynamic model is used to calculate the distribution behavior of rubidium between minerals and fluids and to invert the temperature and pressure conditions of fluid activity; the open-system Rb-Sr isotope mass balance model is used to quantitatively calculate the rubidium inflow / outflow flux and the degree of Sr isotope exchange in each fluid activity; and the rubidium-like isomorphic substitution limit model is used to determine the critical geochemical conditions for rubidium super-enrichment.

9. The comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology according to claim 1, characterized in that: The attention-based temporal graph neural network prediction model takes a mineral generation dataset sorted by absolute time as input. The feature variables include rubidium content, Rb / K ratio, initial 87Sr / 86Sr values, rare earth element characteristic parameters, temperature and pressure conditions, and Sr-Nd isotope source region parameters. The target variables are rubidium enrichment coefficient and fluid migration flux. The model captures the spatial structure correlation of minerals in different generations through graph neural networks and captures the key periods and key ore-controlling parameters that contribute the most to rubidium enrichment through attention mechanisms.

10. The comprehensive investigation system for rubidium enrichment mechanisms in metamorphic rocks based on micro-area in-situ technology according to claim 1, characterized in that: The three-dimensional mineralization prediction and visualization output unit generates a three-dimensional spatial rubidium enrichment potential score based on a machine learning model, combined with a three-dimensional spatial multi-parameter fusion data volume, a three-dimensional spatial rubidium enrichment favorable zone prediction map, a rubidium content -87Sr / 86Sr-time co-evolution map, a spatiotemporal distribution map of multiple phases of fluid activity and rubidium migration paths, and outputs a comprehensive interpretation report on the rubidium enrichment mechanism of metamorphic rocks.

Images