Method for high spatial resolution in situ imaging of molecular sieve carbon deposition deactivation

Abstract

The application discloses a method for high spatial resolution in-situ imaging of carbon deposition deactivation of a molecular sieve, and belongs to the technical field of catalytic material characterization, and aims to solve the problems that the existing in-situ STEM imaging technology is difficult to capture the transient process of carbon deposition precursors, is easy to cause electron damage of the molecular sieve skeleton, and lacks accurate correlation between micro-imaging and macro-product data. A multi-field coupling experimental platform integrating an in-situ atmosphere system, a high-resolution STEM and an iDPC imaging module is built, a multi-component gas pulse type introduction and a rapid switching unit are provided, an electron damage threshold benchmark database is established through system calibration, an atomic-level real space image is collected by using an ultra-low electron dose adaptive imaging mode, and gas phase product information is synchronously collected to realize time sequence correlation. The application can accurately track the molecular transfer and carbon deposition evolution path in a limited space, and provides a reliable technical means for comprehensively analyzing the carbon deposition deactivation mechanism of the molecular sieve and guiding the optimized design of a catalyst.

Description

Technical Field

[0001] This invention relates to the field of catalytic material characterization technology, specifically a high spatiotemporal resolution in-situ imaging method for carbon deposition and deactivation of molecular sieves. Background Technology

[0002] Molecular sieve catalysts are widely used in industrial catalytic processes such as petrochemicals, coal chemicals, and carbon dioxide conversion. Carbon deposition and deactivation are core issues affecting catalytic efficiency and lifespan. In-situ observation of the carbon deposition and deactivation process is a key technique for revealing the deactivation mechanism and optimizing catalyst design. Currently, scanning transmission electron microscopy has become an important tool for studying the microstructure of molecular sieves, and combined with an in-situ atmosphere system, it enables dynamic observation under reaction conditions.

[0003] In existing technologies, in-situ STEM imaging systems typically employ continuous gas introduction, making it difficult to rapidly switch reaction atmospheres and simulate pulsed reaction processes. This results in an inability to accurately capture the transient processes of carbon precursor formation. Furthermore, existing imaging methods use fixed electron dose parameters for data acquisition, failing to differentiate the electronic damage characteristics of molecular sieves with different topologies. This can easily cause irreversible damage to the molecular sieve framework during imaging, making it impossible to obtain atomic-level structural information under real-world conditions. In addition, existing technologies often acquire imaging data or gaseous product data separately, lacking a precise temporal correlation between the two, and thus failing to establish a correspondence between molecular behavior within a confined space and macroscopic reaction products. Summary of the Invention

[0004] The purpose of this invention is to address the aforementioned deficiencies of existing in-situ imaging techniques for carbon deposition deactivation in molecular sieves by providing a high spatiotemporal resolution in-situ imaging method for carbon deposition deactivation in molecular sieves. This method achieves rapid switching and pulsed supply of the reaction atmosphere by improving the gas introduction method of the in-situ atmosphere system. It avoids electronic damage to the molecular sieve framework by establishing an electronic damage threshold benchmark database and adopting an adaptive imaging mode. By achieving precise temporal correlation between imaging data and gas-phase product data, it establishes a correspondence between microscopic molecular behavior and macroscopic reaction products, thereby accurately capturing the transient process of carbon deposition precursor formation and obtaining atomic-level structural information of molecular sieves and confined molecules under real-world conditions. This provides a reliable technical means for comprehensively analyzing the carbon deposition deactivation mechanism of molecular sieves.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a high spatiotemporal resolution in-situ imaging method for deactivation of carbon deposits in molecular sieves, comprising the following steps:

[0006] S1 establishes a multi-field coupled experimental platform integrating an in-situ atmosphere system, a high-resolution scanning transmission electron microscope (STEM), and an integrated differential phase contrast (iDPC) imaging module. The in-situ atmosphere system has independent control capabilities for temperature, pressure, and atmosphere composition, and is equipped with a pulsed introduction and rapid switching unit for multi-component gases.

[0007] This unit controls the on / off timing of different gas passages through a high-speed solenoid valve array, achieving millisecond-level atmosphere switching and pulsed gas supply. It can simulate the rapid reaction kinetics in industrial catalysis, avoiding the averaging of transient processes caused by traditional continuous gas introduction, thus capturing the critical moment of carbon precursor formation.

[0008] S2 performs system calibration on the experimental platform, including electron beam dose calibration, in-situ atmosphere temperature field uniformity calibration, and detector response consistency calibration, and determines the electronic damage threshold benchmark database for different molecular sieve materials.

[0009] This database records the critical dose values ​​at which the framework structure of molecular sieves undergoes irreversible damage by conducting gradient electron dose irradiation experiments on molecular sieves with different topologies and silicon-to-aluminum ratios. This provides a scientific basis for the selection of subsequent imaging parameters and avoids excessive sample damage or insufficient imaging quality caused by empirical parameter settings.

[0010] The S3 places the molecular sieve sample to be tested in the in-situ sample holder, introduces the target reaction atmosphere and adjusts it to the preset operating conditions. It adopts an ultra-low electron dose adaptive imaging mode, combined with a pixel array detector (EMPAD) and a direct-reading camera (K3), to acquire atomic-level real-space images of the molecular sieve framework structure and confined guest molecules in real time.

[0011] This imaging mode dynamically judges the degree of damage to the sample by monitoring the changes in the electron diffraction signal of the sample in real time during the imaging process, thereby automatically adjusting the scanning parameters of the electron beam. Under the premise of ensuring atomic-level imaging resolution, the electron dose is controlled below the sample damage threshold, ensuring that the obtained structural information can truly reflect the intrinsic state of the molecular sieve under working conditions.

[0012] S4 synchronously acquires gaseous product information during the reaction process, and correlates real-space imaging data with gaseous product data in time sequence to track the discrete transport process of molecules and carbon deposition evolution path within the confined space.

[0013] By using a unified clock signal to synchronize the acquisition and triggering of the imaging system and the product analysis system, it is ensured that each frame of image is precisely aligned with the corresponding gas phase product data on the time axis. This enables accurate tracking of the complete process of molecules from adsorption, diffusion, reaction to product desorption, establishing a direct link between microscopic molecular behavior and macroscopic reaction results.

[0014] Preferably, in step S3, while acquiring the real-space image, a femtosecond laser and electron diffraction combined module is introduced to acquire the reciprocal space diffraction signal at the molecular sieve interface, thereby achieving synchronous matching of atomic-level spatial resolution and femtosecond-level temporal resolution, and capturing the electronic transitions of confined molecules and the dynamic vibration process of framework atoms.

[0015] This module employs pump-probe technology, which uses femtosecond laser pulses to excite transient structural changes in the sample and then uses a synchronous pulsed electron beam for diffraction detection. This allows it to capture ultrafast dynamic processes that cannot be observed by traditional continuous imaging techniques, revealing the electronic interactions and energy transfer mechanisms between confined molecules and the molecular sieve framework.

[0016] Preferably, the in-situ atmosphere system in step S1 is further equipped with a mass flow controller array and a gradient concentration mixing unit, which can realize the in-situ simulation of the continuous gradient concentration ratio and competitive adsorption process of at least two guest molecules, wherein the guest molecules include at least one of water, carbon dioxide, oxygen, benzene, and pyridine.

[0017] By controlling the flow rates of different gases using multiple high-precision mass flow controllers, and then achieving uniform mixing through a static mixer, the concentration ratio of the mixed gases can be continuously adjusted to simulate the multi-component competitive adsorption process under different reaction conditions, providing controllable experimental conditions for studying the host-guest interactions in complex reaction systems.

[0018] Preferably, in step S3, the raw signal acquired by the pixel array detector (EMPAD) is processed by a single-molecule-level background noise suppression algorithm. The algorithm extracts single-molecule adsorption and diffusion signals with a signal-to-noise ratio of less than 3 by combining multi-frame superposition and differential filtering.

[0019] The algorithm first performs subpixel-level precise alignment on multiple consecutive images, then removes static background noise through differential operations, and further suppresses random noise using adaptive Gaussian filtering technology. It can extract weak signals at the single-molecule level from raw data with extremely low signal-to-noise ratio, enabling the observation of adsorption sites and diffusion trajectories of individual guest molecules.

[0020] Preferably, in step S4, the atomic-level structural features of the carbon deposition precursor are identified by inter-frame difference analysis of continuous time-series imaging data, and the initial active sites for carbon deposition nucleation and the evolution path of pore blockage are located, distinguishing the carbon deposition type and growth rate in different topological pores.

[0021] By calculating the pixel difference matrix between two adjacent frames, the regions in the image that have changed can be identified, thereby locating the formation location and growth direction of carbon deposit precursors. This allows for the differentiation of the nucleation rate and growth mode of carbon deposits in different topological structures such as straight channels and intersecting channels, providing direct experimental evidence for revealing the channel dependence of carbon deposit deactivation.

[0022] Preferably, the method further includes step S5: inputting the acquired real-space imaging data into the density functional theory (DFT) and molecular dynamics (MD) simulation models, iteratively correcting the intermolecular interaction potential and the electron density distribution of the pore walls in the simulation parameters in real time, and verifying the diffusion path and reaction energy barrier of the confined molecules.

[0023] By using the molecular diffusion trajectory and atomic-level structure information obtained from experiments as constraints, and adjusting the key parameters in the theoretical simulation model through a global optimization algorithm, the simulation results are highly consistent with the experimental results, thereby improving the prediction accuracy of the theoretical simulation and providing reliable theoretical guidance for the rational design of molecular sieve catalysts.

[0024] Preferably, in step S2, based on a pre-established electronic damage threshold benchmark database, the optimal imaging parameters of different topological molecular sieves such as SAPO, ZSM-5, Y, and ZEO are automatically matched. The optimal imaging parameters include electron beam current, scanning speed, and number of frame stacks.

[0025] The system can automatically retrieve the corresponding electronic damage threshold and optimal imaging parameter combination from the database based on the molecular sieve type input by the user, without the need for manual adjustment. This not only improves experimental efficiency but also ensures the consistency and comparability of experimental results from different batches.

[0026] Preferably, in step S4, the gas phase product information is acquired by an online gas chromatography-mass spectrometry (GC-MS) system, and the GC-MS and STEM imaging system achieve millisecond-level time synchronization through a unified timing triggering unit to establish a one-to-one correspondence between "adsorbed molecules and gas phase products".

[0027] This timing triggering unit can simultaneously send high-precision synchronization trigger signals to the STEM imaging system and the GC-MS system, ensuring that the acquisition time start points of the two are completely consistent. It can also flexibly adjust the acquisition frequency according to experimental needs to achieve millisecond-level time synchronization accuracy, thereby accurately establishing the correspondence between the adsorbed molecular structure and the composition of gaseous products.

[0028] Preferably, the ultra-low electron dose adaptive imaging mode in step S3 specifically involves: dynamically adjusting the scanning step size and dwell time of the electron beam according to the real-time structural damage degree of the molecular sieve sample, so as to control the electron dose of a single frame imaging to below 30% of the corresponding molecular sieve electron damage threshold.

[0029] When the system detects attenuation of the electron diffraction signal of the sample, it automatically reduces the residence time of the electron beam at each pixel or increases the scanning step size to reduce the accumulation of electron dose in local areas, while maintaining high imaging resolution in undamaged areas, thus achieving the best balance between sample damage control and imaging quality.

[0030] Preferably, in step S4, the real-space imaging data is reconstructed in three dimensions to quantify the bond lengths and bond energies of van der Waals interactions, hydrogen bonding interactions, and proton transfer interactions between the host and guest, and to establish a correlation map between the active sites of the molecular sieve and the carbon deposition deactivation rate.

[0031] By reconstructing a series of atomic-level images acquired from different angles in three dimensions, a three-dimensional atomic model of molecular sieve channels and confined molecules is obtained. Then, combined with quantum chemical calculation methods, the strength of various interactions between the host and guest is quantitatively calculated, thereby establishing a quantitative structure-activity relationship between the structure of molecular sieve active sites and the carbon deposition deactivation rate.

[0032] Compared with existing technologies, this invention provides a high spatiotemporal resolution in-situ imaging method for deactivation of carbon deposits in molecular sieves, which has the following beneficial effects:

[0033] This invention, by equipping an in-situ atmosphere system with a pulsed introduction and rapid switching unit for multi-component gases, enables rapid switching and pulsed supply of the reaction atmosphere, accurately capturing the transient process of carbon deposit precursor formation, and providing technical support for analyzing the initial stage of carbon deposit deactivation.

[0034] This invention calibrates the experimental platform and establishes a benchmark database of electronic damage thresholds for different molecular sieve materials. Combined with an ultra-low electron dose adaptive imaging mode, it can dynamically adjust imaging parameters according to the structural characteristics of the molecular sieve. This ensures high-resolution imaging while avoiding electronic damage to the molecular sieve framework structure, thus obtaining atomic-level structural information under real-world conditions.

[0035] This invention, by synchronously acquiring gaseous product information during the reaction process and temporally correlating real-space imaging data with gaseous product data, can establish the correspondence between molecular discrete transport process, carbon deposition evolution path and macroscopic reaction products within a confined space, providing complete data support for a comprehensive analysis of the carbon deposition and deactivation mechanism of molecular sieves. Attached Figure Description

[0036] Figure 1 This diagram illustrates the steps of the high spatiotemporal resolution in-situ imaging method for deactivation of carbon deposits in molecular sieves according to the present invention.

[0037] Figure 2 This is a characterization diagram of the distribution of shell-type carbon deposits and the evolution of pore-confined carbon deposits in ZSM-5 molecular sieves during methanol conversion. Detailed Implementation

[0038] Please see Figure 1-2 This invention provides a technical solution: a high spatiotemporal resolution in-situ imaging method for deactivation of carbon deposits in molecular sieves, which is implemented sequentially according to the following steps:

[0039] S1 establishes a multi-field coupled experimental platform integrating an in-situ atmosphere system, a high-resolution scanning transmission electron microscope (STEM), and an integrated differential phase contrast (iDPC) imaging module. The in-situ atmosphere system possesses independent controllability of temperature, pressure, and atmosphere composition, and is equipped with a pulsed introduction and rapid switching unit for multi-component gases. In the specific implementation of this step, a commercial high-resolution scanning transmission electron microscope is first used as the basic imaging platform. Gas inlet and outlet pipes for the in-situ atmosphere system are installed at the vacuum interface reserved on the side of its sample chamber. The pipes are made of stainless steel and have their inner walls polished to reduce gas adsorption and residue. The core control unit of the in-situ atmosphere system is independent of the main control system of the electron microscope, and interacts with the control computer of the electron microscope via the RS-485 communication protocol to achieve linked control of imaging parameters and atmosphere parameters. The temperature control module of the in-situ atmosphere system employs a combination of resistance heating and liquid nitrogen cooling. The heating element is a platinum-rhodium alloy wire embedded inside the sample stage of the in-situ sample rod. The cooling circuit is connected to an external liquid nitrogen Dewar via flexible metal tubing. The temperature control range covers room temperature to the preset maximum reaction temperature, and temperature control accuracy is achieved using a proportional-integral-derivative (PID) algorithm. The pressure control module consists of a high-precision back pressure valve, a vacuum gauge, and a molecular pump. The back pressure valve is installed on the gas outlet line, the vacuum gauge monitors the atmosphere pressure inside the sample chamber in real time, and the molecular pump maintains the system's basic vacuum level. The pressure control range covers high vacuum to the preset maximum reaction pressure.

[0040] The multi-component gas pulsed introduction and rapid switching unit consists of a high-speed solenoid valve array, a gas buffer chamber, and a timing controller. The high-speed solenoid valves employ a direct-acting structure with a response time no greater than a preset switching time threshold. Each solenoid valve corresponds to an independent gas path, connected to different gas cylinders. The gas buffer chamber is located between the solenoid valves and the sample chamber to stabilize the pressure and flow rate of the pulsed gas, preventing gas pressure fluctuations from interfering with the imaging system. The timing controller uses a programmable logic controller (PLC) as its core, capable of precisely controlling the opening and closing times of each solenoid valve, as well as the switching sequence between different solenoid valves, according to a preset program, achieving pulsed supply of single-component gases and rapid switching of multi-component gases. A gas purification unit, including a molecular sieve dryer and a deoxygenator, is also installed in the gas introduction pipeline to remove moisture and oxygen impurities from the gas, preventing contamination and interference from impurity gases on the molecular sieve sample.

[0041] The integrated differential phase contrast (iDPC) imaging module is integrated into the detector system of the electron microscope, consisting of a four-quadrant ring detector and a dedicated signal processing circuit. The four-quadrant ring detector is mounted below the electron microscope's projection lens to collect scattered electrons transmitted through the sample. The signal processing circuit performs real-time acquisition and differential calculations on the electron signals from the four quadrants to generate iDPC image signals. The iDPC imaging module works synchronously with the electron microscope's scanning system; the deflection signals of the scanning coils are simultaneously input into the iDPC signal processing circuit to ensure a one-to-one correspondence between image pixels and scanning positions. After the platform is built, the system's vacuum tightness needs to be tested. The sample chamber is evacuated to a high vacuum, all valves are closed, and after a preset time, the change in vacuum level is monitored to ensure that the system's leakage rate meets the requirements of in-situ experiments.

[0042] S2 performs system calibration on the experimental platform, including electron beam dose calibration, in-situ atmosphere temperature field uniformity calibration, and detector response consistency calibration, establishing a benchmark database of electron damage thresholds for different molecular sieve materials. In the specific implementation of electron beam dose calibration, a Faraday cup is used as the standard dose measurement device. The Faraday cup is installed at the sample stage position of the in-situ sample rod, replacing the molecular sieve sample for electron beam dose measurement. The electron beam emitted by the electron gun of the electron microscope is focused by the condenser system and irradiates the Faraday cup. The electron current collected by the Faraday cup is measured by a high-precision picoammeter, recording the electron dose values ​​under different electron beam currents, scanning speeds, and scanning areas, establishing a calibration curve between electron beam parameters and electron dose. During calibration, the accelerating voltage of the electron microscope needs to be kept constant, and different condenser aperture sizes and probe currents are calibrated to ensure that the measurement accuracy of the electron beam dose meets the requirements across the entire imaging parameter range.

[0043] In-situ atmosphere temperature field uniformity calibration employs a thermocouple array method. Multiple miniature thermocouples are evenly distributed on the sample stage surface, with the measuring ends of the thermocouples in close contact with the sample stage surface, and the other ends connected to an external temperature recorder via wires. Inert gas is introduced into the in-situ atmosphere system and adjusted to a preset pressure. The sample stage temperature is then gradually increased to different set values, and the measurements of each thermocouple at different temperatures are recorded to plot the temperature distribution curve on the sample stage surface. Based on the temperature distribution curve, the power distribution of the heating element and the PID control parameters are adjusted to ensure the temperature uniformity of the sample stage surface meets experimental requirements. Temperature field uniformity calibration needs to be performed separately for different atmosphere compositions and pressure conditions because the thermal conductivity of the gas affects the temperature distribution of the sample stage.

[0044] Detector response consistency calibration includes calibration of the pixel array detector (EMPAD) and the direct-reading camera (K3). For the pixel array detector, a uniform electron beam illuminates the entire photosensitive surface of the detector, the response signal of each pixel is recorded, and then the average response value of all pixels is calculated. The response signal of each pixel is normalized to generate the detector's response correction matrix. For the direct-reading camera, calibration is performed using a standard grayscale chart. The standard grayscale chart is placed within the camera's field of view, and grayscale images are captured at different exposure times. A calibration curve between exposure time and grayscale value is established to ensure that the camera's grayscale response linearity meets the requirements.

[0045] The establishment of the electronic damage threshold benchmark database was accomplished through gradient electron dose irradiation experiments. First, representative molecular sieve samples with different topologies, silicon-to-aluminum ratios, and modification methods were selected and prepared into thin samples suitable for electron microscopy observation. The samples were mounted on in-situ sample holders and evacuated to high vacuum. Then, an electron beam was focused onto a selected region of the sample, and the electron dose was gradually increased according to a preset gradient. After each dose, high-resolution TEM images and electron diffraction patterns of the samples were acquired to observe changes in the sample's skeletal structure. When the electron diffraction pattern of the sample showed obvious diffuse scattering or the disappearance of characteristic diffraction peaks, or when an amorphous region appeared in the high-resolution TEM image, the electron dose value at this point was recorded as the electronic damage threshold of that molecular sieve sample. The electronic damage threshold data of all tested samples, along with the corresponding sample structural parameters and imaging parameters, were compiled and stored to form the electronic damage threshold benchmark database. The database adopts a relational database structure, including a sample information table, a damage threshold data table, and an imaging parameter table, supporting retrieval and querying based on sample type, structural parameters, and other conditions.

[0046] In step S3, the molecular sieve sample to be tested is placed in the in-situ sample holder, the target reaction atmosphere is introduced, and the conditions are adjusted to the preset operating conditions. Using an ultra-low electron dose adaptive imaging mode, combined with a pixel array detector (EMPAD) and a direct-reading camera (K3), atomic-level real-space images of the molecular sieve framework structure and confined guest molecules are acquired in real time. Specifically, the molecular sieve sample to be tested is first ground into powder, then dispersed on a microgrid covered with a carbon support film. Excess solvent is absorbed with filter paper and the sample is then dried. The prepared sample is mounted on the sample stage of the in-situ heated sample holder, ensuring good thermal contact between the sample and the stage. The in-situ sample holder is then slowly inserted into the sample chamber of the electron microscope, maintaining a stable vacuum level in the sample chamber during insertion to prevent sample contamination.

[0047] After the vacuum level in the sample chamber returns to the preset value, the in-situ atmosphere system is activated. First, inert gas is introduced to purge the pipelines and sample chamber, removing residual air and impurities. Then, the target reaction atmosphere is introduced according to the preset procedure, and the flow rate of the mass flow controller is adjusted to the set value. Once the gas flow rate stabilizes, the back pressure valve is adjusted to raise the pressure in the sample chamber to the preset reaction pressure. Simultaneously, the heating system is activated, raising the temperature of the sample stage to the preset reaction temperature according to the preset heating rate, and maintaining this temperature for a preset time to allow the sample to reach thermal equilibrium. During atmosphere and temperature regulation, the pressure and temperature changes in the sample chamber are monitored in real time to ensure stable operating conditions.

[0048] The specific implementation process of the ultra-low electron dose adaptive imaging mode is as follows: First, based on the type of molecular sieve sample to be tested, the corresponding electron damage threshold and recommended initial imaging parameters, including electron beam current, scanning speed, number of frame stacks, and detector gain, are retrieved from the electron damage threshold benchmark database. Then, the electron beam is focused on a selected area of ​​the sample, and initial electron diffraction patterns and low-magnification images are acquired to confirm the sample's position and orientation. The adaptive imaging control program is started, and the program acquires the sample's electron diffraction signals in real time. The intensity of characteristic peaks and the full width at half maximum (FWHM) variation of the diffraction pattern are analyzed through Fast Fourier Transform (FFT) to determine the degree of structural damage to the sample. When the intensity of the diffraction characteristic peaks drops to a preset damage warning threshold, the program automatically adjusts the electron beam scanning parameters to reduce the electron dose accumulation rate; when the intensity of the diffraction characteristic peaks returns to normal levels, the program appropriately increases the scanning parameters to obtain higher imaging quality.

[0049] The pixel array detector (EMPAD) and the direct-reading camera (K3) operate synchronously. The EMPAD is used to acquire high-resolution iDPC images, while the direct-reading camera is used to acquire low-magnification bright-field images and electron diffraction patterns. The acquisition trigger signals for both detectors are synchronously output by the electron microscope's scanning system, ensuring precise temporal correspondence between the iDPC images and electron diffraction patterns. The acquired raw image data is transmitted in real-time to the control computer for storage and preliminary processing, including dark current subtraction, gain correction, and noise filtering. During imaging, the program automatically records the acquisition time, electron dose, imaging parameters, and corresponding operating conditions for each frame, providing complete metadata for subsequent data analysis.

[0050] S4 synchronously acquires gaseous product information during the reaction process, and correlates the real-space imaging data with the gaseous product data in time to track the discrete transport process of molecules and the carbon deposition evolution path within the confined space. In this step, gaseous product information is acquired using an online gas chromatography-mass spectrometry (GC-MS) system. The GC-MS inlet is connected to the gas outlet of the in-situ atmosphere system via a heated transfer line. The temperature of the transfer line is maintained above a preset value to prevent condensation of gaseous products within the line. A split valve is installed in the transfer line to introduce a portion of the outlet gas into the GC-MS for analysis, while the other portion is discharged from the system through a back pressure valve. The split ratio is adjusted according to the concentration of the reaction products.

[0051] The GC-MS and STEM imaging systems achieve millisecond-level time synchronization through a unified timing trigger unit. This unit uses a highly stable temperature-controlled crystal oscillator as its clock source to generate a high-precision synchronization trigger signal, which is simultaneously sent to the control computer of the STEM imaging system and the control system of the GC-MS. When the experiment begins, the timing trigger unit sends a start trigger signal, and both the STEM imaging system and GC-MS simultaneously begin data acquisition, assigning a precise timestamp to each acquired data point. The timestamp accuracy reaches the millisecond level, ensuring precise alignment of the imaging data and gaseous product data on the timeline.

[0052] During data acquisition, GC-MS samples at preset time intervals, with each sampling cycle comprising four stages: sample introduction, separation, detection, and data processing. The STEM imaging system continuously acquires images at a preset frame rate, with each frame corresponding to a unique timestamp. After acquisition, all imaging data and gaseous product data are imported into dedicated data analysis software, and the two are correlated temporally based on their timestamps. For the gaseous product composition data at each time point, the corresponding imaging data frame is located, and the correspondence between the molecular adsorption, diffusion, and reaction behavior within the molecular sieve channels at that time point and the gaseous product composition is analyzed.

[0053] By employing inter-frame difference analysis of continuous temporal imaging data, the atomic-level structural features of carbon deposition precursors are identified, and the initial active sites for carbon deposition nucleation and the evolutionary path of pore blockage are located. The specific process of inter-frame difference analysis is as follows: First, multiple consecutively acquired images are precisely aligned at the sub-pixel level to eliminate image displacement caused by sample drift and vibration. Then, the pixel difference matrix between adjacent images is calculated. Pixels with larger absolute values ​​in the difference matrix correspond to regions in the image where changes occur; these regions represent the locations where molecular adsorption, diffusion, or reactions take place. By analyzing a series of difference matrices, the molecular trajectory and the growth process of carbon deposition precursors can be tracked, determining the initial location of carbon deposition nucleation and the evolutionary direction of pore blockage. Simultaneously, by combining the pore characteristics of molecular sieves with different topologies, the carbon deposition growth patterns and rate differences in different regions such as straight channels, sinusoidal channels, and intersecting channels are distinguished.

[0054] S5 inputs the acquired real-space imaging data into density functional theory (DFT) and molecular dynamics (MD) simulation models, iteratively correcting the intermolecular interaction potential and pore wall electron density distribution in the simulation parameters in real time, and verifying the diffusion path and reaction energy barrier of confined molecules. In the specific implementation of this step, firstly, based on the structural parameters of the molecular sieve sample used in the experiment, a corresponding atomic-level structural model is constructed, including the framework structure of the molecular sieve, the distribution of acidic sites, and the geometric features of the pores. Then, the adsorption sites and structural information of the confined molecules observed in the experiment are used as initial conditions and input into the DFT calculation model. Appropriate functionals and basis sets are used for geometric optimization and electronic structure calculation to obtain the interaction energy and electron density distribution between the host and guest molecules.

[0055] The interaction energy parameters obtained from DFT calculations were substituted into the molecular dynamics simulation model to construct a molecular dynamics simulation system. This system included a molecular sieve framework, confined guest molecules, and reaction atmosphere molecules. Molecular dynamics simulations were performed using either a canonical ensemble (NVT) or a grand canonical ensemble (GCMC), with simulation temperatures and pressures consistent with experimental conditions. During the simulation, the molecular trajectories, diffusion coefficients, and probabilities of reaction events were recorded to obtain the diffusion paths and reaction energy barriers of the confined molecules within the molecular sieve channels.

[0056] The simulated molecular diffusion trajectories and reaction energy barriers are compared with experimentally observed results. If a significant deviation exists, a global optimization algorithm is used to iteratively correct the intermolecular interaction potentials and pore wall electron density distributions in the simulation parameters. The corrected parameters are then re-introduced into the simulation model for calculation until the deviation between the simulation and experimental results is within a preset allowable range. This method, combining experimental and theoretical simulation, not only verifies the correctness of experimental observations but also reveals the microscopic nature of host-guest interactions, providing theoretical guidance for the rational design of molecular sieve catalysts.

[0057] During the implementation of the above steps, a femtosecond laser and electron diffraction co-processing module can be introduced as needed to collect reciprocal space diffraction signals at the molecular sieve interface, achieving synchronous matching between atomic-level spatial resolution and femtosecond-level temporal resolution. The femtosecond laser and electron diffraction co-processing module adopts a pump-probe architecture. The laser pulse output from the femtosecond laser system is divided into two beams: one beam is used as the pump light, focused onto the sample to excite transient structural changes, and the other beam is used as the synchrotron light to trigger the electron gun to generate a pulsed electron beam. The pulsed electron beam is focused onto the sample by a condenser lens system, and the resulting electron diffraction signal is collected by the detector. By precisely adjusting the time delay between the pump light and the detector electron beam, electron diffraction patterns at different delay times can be collected, thereby obtaining the evolution of the sample structure over time and capturing the electronic transitions of confined molecules and the dynamic vibrations of framework atoms.

[0058] The in-situ atmosphere system is also equipped with a mass flow controller array and a gradient concentration mixing unit, enabling in-situ simulation of continuous gradient concentration ratios and competitive adsorption processes for at least two guest molecules. The mass flow controller array consists of multiple high-precision mass flow controllers, each corresponding to a gas path, allowing for precise control of gas flow. The gradient concentration mixing unit comprises a static mixer and a concentration sensor. After thorough mixing of gases at different flow rates in the static mixer, the concentration of the mixed gas is monitored in real time by the concentration sensor. By programmatically controlling the flow ratios of different mass flow controllers, continuous linear changes in the mixed gas concentration can be achieved, simulating atmospheric composition changes at different reaction stages and studying the competitive adsorption and reaction behavior between multi-component gases.

[0059] The raw signals acquired by the pixel array detector (EMPAD) are processed by a single-molecule-level background noise suppression algorithm. This algorithm extracts single-molecule adsorption and diffusion signals with a signal-to-noise ratio (SNR) below 3 by combining multi-frame superposition and differential filtering. The specific implementation steps of the algorithm are as follows: First, the continuously acquired multi-frame raw images are precisely aligned at the sub-pixel level to eliminate image displacement caused by sample drift and vibration. Then, the aligned images are superimposed in multiple frames to improve the overall SNR of the image. Next, background estimation is performed on the superimposed image, and the static background distribution of the image is obtained by using a polynomial fitting method. The static background is subtracted from the original image to obtain a differential image containing the signal and random noise. Finally, adaptive Gaussian filtering is applied to the differential image, automatically adjusting the size and standard deviation of the filtering window according to the local noise level of the image, suppressing random noise while retaining weak single-molecule-level signals.

[0060] Based on a pre-established electronic damage threshold benchmark database, the system automatically matches the optimal imaging parameters for different topological structures of SAPO, ZSM-5, Y, and ZEO series molecular sieves. According to user-input information such as the type of molecular sieve sample, silica-to-alumina ratio, and modification method, the system automatically searches the electronic damage threshold benchmark database to find the corresponding electronic damage threshold and the experimentally validated optimal combination of imaging parameters, including electron beam current, scanning speed, number of frame stacks, detector gain, and exposure time. The system automatically loads these parameters into the electron microscope's control program, eliminating the need for tedious manual parameter adjustments. This not only improves experimental efficiency but also ensures the consistency and comparability of results from different batches of experiments.

[0061] Three-dimensional reconstruction of real-space imaging data was performed to quantify the bond lengths and bond energies of van der Waals interactions, hydrogen bonds, and proton transfer interactions between the host and guest molecules, establishing a correlation map between the active sites of the molecular sieve and the rate of carbon deposition and deactivation. Three-dimensional reconstruction employed electron tomography, tilting the sample stage at different angles to acquire a series of atomic-level images from various perspectives. Weighted back projection or iterative reconstruction algorithms were then used to reconstruct these images in three dimensions, yielding the three-dimensional atomic density distribution of the molecular sieve channels and confined molecules. Geometric parameters such as bond lengths and bond angles between the host and guest molecules were extracted from the three-dimensional reconstruction results. Combined with quantum chemical calculations, the strengths of van der Waals interactions, hydrogen bonds, and proton transfer interactions were quantitatively calculated. Through statistical analysis of a large amount of experimental data, a quantitative structure-activity relationship was established between the structural parameters of the active sites of the molecular sieve, the strength of host-guest interactions, and the rate of carbon deposition and deactivation, forming a correlation map. This provides a direct theoretical basis for the anti-carbon deposition modification of molecular sieve catalysts.

[0062] Through the above-described specific embodiments, this invention achieves high spatiotemporal resolution in-situ imaging of the carbon deposition and deactivation process of molecular sieves, solving technical problems in the prior art such as the inability to capture transient processes, the susceptibility of samples to electronic damage, and the lack of correlation between microscopic and macroscopic data. It provides a powerful technical means for deeply revealing the microscopic mechanism of carbon deposition and deactivation of molecular sieves.

[0063] The present invention will be described below based on specific implementation examples:

[0064] This embodiment takes the methanol conversion reaction as an example, uses ZSM-5 molecular sieve as a catalyst, and employs the high spatiotemporal resolution in-situ imaging method for carbon deposition and deactivation of molecular sieves described in this invention to observe the carbon deposition formation and distribution characteristics of ZSM-5 molecular sieve during the methanol conversion process.

[0065] HZSM-5 molecular sieve powder was dispersed in anhydrous ethanol and ultrasonically dispersed for 5–15 min to obtain a uniform suspension. The suspension was then dropped onto an in-situ atmosphere transmission electron microscope chip or a high-temperature resistant sample grid and dried to obtain the sample to be tested. The sample to be tested was pretreated in an inert atmosphere at 350–500 °C for 0.5–2 h to remove adsorbed water and surface impurities.

[0066] The pretreated sample was placed in an in-situ atmosphere STEM sample holder and mounted into a scanning transmission electron microscope. Before the reaction, initial images of ZSM-5 molecular sieves were acquired under an inert atmosphere using low-dose HAADF-STEM and iDPC-STEM modes, recording their particle morphology, pore structure, and outer surface condition. Based on a pre-established electronic damage threshold benchmark database, the safe electron dose range for ZSM-5 molecular sieves was determined, and the electron dose was controlled below this safe threshold during subsequent imaging.

[0067] The in-situ reaction tank was heated to 400–450℃, and methanol vapor was introduced using helium or nitrogen as the carrier gas. The methanol partial pressure was 0.5–10 kPa, and the total gas flow rate was 5–50 mL / min. Methanol was fed continuously or in a pulsed manner. During the reaction, HAADF-STEM images, iDPC-STEM images, C / O / Si elemental distribution images, and outlet gas phase product signals of the ZSM-5 molecular sieve were simultaneously acquired, and the image acquisition time and product detection time were synchronously correlated.

[0068] In the initial stage of the reaction, the ZSM-5 molecular sieve particles maintained a complete crystal morphology, and no obvious carbon deposition was observed in the pore region. As the reaction proceeded, contrast-enhanced regions appeared in the pore intersection areas, grain edges, and near the outer surface of the ZSM-5 molecular sieve, indicating that carbon precursors began to form and gradually accumulate. As the reaction continued, carbonaceous species accumulated further, forming localized carbon deposits inside the pores and layered carbon deposits on the outer surface of the particles.

[0069] After 96 hours of reaction, elemental distribution analysis was performed on the ZSM-5 molecular sieve particles. The results showed that Si and O signals were mainly distributed in the molecular sieve framework region, while C signals were significantly enriched at the particle edges and near-surface regions, indicating that the carbon deposits formed during methanol conversion were distributed in a shell-like pattern rather than uniformly throughout the entire particle. Further high-resolution iDPC-STEM observation revealed discrete carbon deposit regions within the pores of the ZSM-5 molecular sieve, with continuous or semi-continuous layered carbonaceous deposits on the outer surface; some of the outer carbon deposit regions showed an interlayer spacing of approximately 0.34 nm, indicating a certain degree of layered ordered structure.

[0070] Meanwhile, continuous imaging characterization of the confined pore region revealed that as the methanol conversion reaction proceeds, the carbon deposit molecules within the pores gradually transform from an initial disordered or weakly ordered state into more structured ordered carbon species, indicating that the confined pore environment can influence the aggregation, rearrangement, and structural evolution of carbon deposit precursors.

[0071] This embodiment illustrates that the method of the present invention can achieve in-situ characterization of the carbon deposition and deactivation process of ZSM-5 molecular sieve under methanol conversion reaction conditions, distinguish between internal and external carbon deposition, identify shell-type carbon deposition distribution, and reveal the dynamic evolution of the pore-confined carbon deposition molecular structure under reaction conditions.

Claims

1. A high spatiotemporal resolution in-situ imaging method for deactivation of carbon deposits in molecular sieves, characterized in that, Includes the following steps: S1. Construct a multi-field coupled experimental platform integrating an in-situ atmosphere system, a high-resolution scanning transmission electron microscope, and an integral differential phase contrast imaging module. The in-situ atmosphere system has independent control capabilities for temperature, pressure, and atmosphere composition, and is equipped with a pulsed introduction and rapid switching unit for multi-component gases. S2. Perform system calibration on the experimental platform, including electron beam dose calibration, in-situ atmosphere temperature field uniformity calibration, and detector response consistency calibration, and determine the electronic damage threshold benchmark database for different molecular sieve materials. S3. Place the molecular sieve sample to be tested in the in-situ sample rod, introduce the target reaction atmosphere and adjust it to the preset working conditions. Use the ultra-low electron dose adaptive imaging mode, combined with the pixel array detector and the direct reading camera, to acquire atomic-level real-space images of the molecular sieve framework structure and confined guest molecules in real time. S4. Simultaneously acquire gaseous product information during the reaction process, correlate real-space imaging data with gaseous product data in time sequence, and track the discrete transport process of molecules and carbon deposition evolution path within the confined space.

2. The high spatiotemporal resolution in-situ imaging method for molecular sieve carbon deposition deactivation according to claim 1, characterized in that, In step S3, while acquiring real-space images, a femtosecond laser and electron diffraction combined module is introduced to acquire reciprocal space diffraction signals at the molecular sieve interface, achieving synchronous matching of atomic-level spatial resolution and femtosecond-level temporal resolution, and capturing the electronic transitions of confined molecules and the dynamic vibration processes of framework atoms.

3. The high spatiotemporal resolution in-situ imaging method for molecular sieve carbon deposition deactivation according to claim 1, characterized in that, The in-situ atmosphere system in step S1 is also equipped with a mass flow controller array and a gradient concentration mixing unit, which can realize the in-situ simulation of continuous gradient concentration ratio and competitive adsorption process of at least two guest molecules, including at least one of water, carbon dioxide, oxygen, benzene, and pyridine.

4. The high spatiotemporal resolution in-situ imaging method for carbon deposition deactivation of molecular sieves according to claim 1, characterized in that, In step S3, the raw signal collected by the pixel array detector is processed by a single-molecule-level background noise suppression algorithm. The algorithm extracts single-molecule adsorption and diffusion signals with a signal-to-noise ratio of less than 3 by combining multi-frame superposition and differential filtering.

5. The high spatiotemporal resolution in-situ imaging method for deactivation of carbon deposits in molecular sieves according to claim 1, characterized in that, In step S4, the atomic-level structural features of carbon deposition precursors are identified through inter-frame differential analysis of continuous time-series imaging data, and the initial active sites for carbon deposition nucleation and the evolution path of pore blockage are located, distinguishing the carbon deposition types and growth rates in different topological pores.

6. The high spatiotemporal resolution in-situ imaging method for deactivation of carbon deposits in molecular sieves according to claim 1, characterized in that, It also includes step S5: inputting the collected real-space imaging data into the density functional theory and molecular dynamics simulation model, iteratively correcting the intermolecular interaction potential and the electron density distribution of the pore walls in the simulation parameters in real time, and verifying the diffusion path and reaction energy barrier of the confined molecules.

7. The high spatiotemporal resolution in-situ imaging method for deactivation of carbon deposits in molecular sieves according to claim 1, characterized in that, In step S2, based on a pre-established electronic damage threshold benchmark database, the optimal imaging parameters for different topological molecular sieves of SAPO, ZSM-5, Y, and ZEO series are automatically matched. The optimal imaging parameters include electron beam current, scanning speed, and number of frame stacks.

8. The high spatiotemporal resolution in-situ imaging method for molecular sieve carbon deposition deactivation according to claim 1, characterized in that, In step S4, gas phase product information is acquired by an online gas chromatograph-mass spectrometer, and the GC-MS and STEM imaging system achieve millisecond-level time synchronization through a unified timing triggering unit, establishing a one-to-one correspondence between adsorbed molecules and gas phase products.

9. The high spatiotemporal resolution in-situ imaging method for carbon deposition deactivation of molecular sieves according to claim 1, characterized in that, The ultra-low electron dose adaptive imaging mode in step S3 specifically involves dynamically adjusting the scanning step size and dwell time of the electron beam based on the real-time structural damage level of the molecular sieve sample, thereby controlling the electron dose of a single frame image to below 30% of the corresponding molecular sieve electron damage threshold.

10. The high spatiotemporal resolution in-situ imaging method for carbon deposition deactivation of molecular sieves according to claim 1, characterized in that, In step S4, the real-space imaging data is reconstructed in three dimensions to quantify the bond lengths and bond energies of van der Waals interactions, hydrogen bonding interactions, and proton transfer interactions between the host and guest, and to establish a correlation map between the active sites of the molecular sieve and the carbon deposition deactivation rate.

Images