Photocatalytic hydrogen production interface reaction dynamic regulation and control method based on in-situ Raman spectrum-molecular dynamics coupling simulation

Abstract

The invention discloses a photocatalytic hydrogen production interface reaction dynamic regulation and control method based on in-situ Raman spectrum-molecular dynamics coupling simulation, and relates to the technical field of photocatalytic hydrogen production. The method comprises the following steps: firstly, building an in-situ Raman system containing a corrosion-resistant light-transmitting reaction tank, monitoring microscopic information such as lattice stress change and TiO2 crystal phase transformation caused by CdS photocorrosion in real time, and capturing characteristic peaks of intermediates such as. OH and the like and O-H bond vibration frequency deviation; constructing a molecular dynamics model containing water molecules on the surface of the catalyst and adsorbed on the surface of the catalyst in combination with spectral data, and obtaining the influence of an IrRu monatomic site on water orientation, M-H bond length evolution and other atomic scale mechanisms through de novo calculation simulation; using a data fusion algorithm to correlate the spectral characteristic peak intensity with the simulation energy change, and locking the reaction speed step; finally, targeted regulation and control are performed, for example, a heterojunction is constructed to inhibit CdS light corrosion, the interlayer spacing of the layered catalyst is adjusted to strengthen charge separation, and the reaction temperature and hydrogen partial pressure are collaboratively optimized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of photocatalytic hydrogen production, and particularly relates to a photocatalytic hydrogen production interface reaction dynamic regulation method based on in-situ Raman spectrum-molecular dynamics coupling simulation. BACKGROUND

[0002] Hydrogen energy, as a clean and efficient secondary energy, has only water as its combustion product, which can effectively alleviate the environmental pollution and energy crisis problems caused by the use of traditional fossil energy. Therefore, photocatalytic hydrogen production technology, as an important way to realize water decomposition for hydrogen production by using solar energy, has attracted widespread attention from the global scientific research community and the industrial community. The core process of photocatalytic hydrogen production reaction occurs at the interface of the catalyst and the reaction medium. The microstructure changes, charge transfer efficiency, and intermediate conversion path at the interface directly determine the hydrogen production efficiency. Therefore, precise regulation of the interface reaction is the key to improving the performance of photocatalytic hydrogen production. The main challenge of current photocatalytic hydrogen production technology is the complexity of the interface reaction mechanism and the difficulty in real-time monitoring, which leads to a lack of targeted regulation methods. Traditional characterization methods such as offline X-ray diffraction and transmission electron microscopy can only obtain static information before and after the reaction, and cannot capture the dynamic changes in the microstructure during the reaction process. For example, key information such as the initial time of photocorrosion of CdS photocatalyst during hydrogen production and the intermediate state structure of TiO2 phase transition is difficult to be effectively captured, making the optimization design of photocatalysts lack direct experimental basis. Although molecular dynamics simulation technology can analyze the reaction mechanism at the atomic scale, such as revealing the interface mechanism by simulating the influence of IrRu monatomic sites on the orientation of water molecules, the accuracy of the simulation results depends on the matching degree with the actual reaction conditions. However, existing simulations are mostly based on idealized models, lacking correlation verification with dynamic data of real reaction systems, which limits the guiding effect of simulation conclusions on actual regulation. At the same time, existing regulation methods mostly rely on empirical parameter adjustment, such as blindly changing the catalyst composition or reaction temperature, which makes it difficult to achieve directional regulation of key steps of the interface reaction, restricting the further improvement of the efficiency of photocatalytic hydrogen production. Therefore, developing a method that can real-time monitor the dynamic changes of the interface, accurately analyze the reaction mechanism, and achieve directional regulation, has become an urgent need to break through the bottleneck of photocatalytic hydrogen production technology, which is also the research and development background and core goal of the present application. SUMMARY

[0003] The application aims to provide a photocatalytic hydrogen production interface reaction dynamic regulation method based on in-situ Raman spectrum-molecular dynamics coupling simulation, which comprehensively solves the problems of real-time monitoring difficulty of interface reaction, lack of precise targeting of regulation measures and low hydrogen production efficiency in existing photocatalytic hydrogen production technology. The method innovatively combines in-situ Raman spectrum technology and molecular dynamics simulation, captures the key information such as catalyst microstructure change, intermediate generation and conversion in the reaction process in real time through in-situ Raman spectrum, analyzes the internal mechanism of the reaction in depth from the atomic scale by means of molecular dynamics simulation, and then builds the correlation bridge between macroscopic spectral information and microscopic atomic motion through coupling analysis of the two, so as to realize the precise and dynamic regulation of photocatalytic hydrogen production interface reaction, thereby greatly improving the efficiency of photocatalytic hydrogen production and providing strong technical support for the practical application of photocatalytic hydrogen production technology.

[0004] In order to achieve the purpose of the application, the technical scheme adopted is: A photocatalytic hydrogen production interface reaction dynamic regulation method based on in-situ Raman spectrum-molecular dynamics coupling simulation, comprising the following steps: S1. Building an in-situ Raman spectrum monitoring system The system comprises a light-transmitting reaction tank with a magnetic drive stirring paddle, a laser light source with a wavelength of 532 nm, a spectrometer and a data acquisition device; Specifically, the system comprises a corrosion-resistant light-transmitting reaction tank with a volume of 50-200 mL, and high-purity quartz glass (purity ≥ 99.9%) is selected as the material of the reaction tank, because high-purity quartz glass not only has excellent chemical stability and can resist corrosion of reaction solutions (such as acidic or alkaline aqueous solutions), but also has a light transmittance of more than 90% in the Raman detection range of 200-4000 cm -1 The light transmittance of the reaction tank body is equipped with a gas inlet and a gas outlet with an inner diameter of 3-5 mm, which can ensure smooth gas flow and reduce the disturbance of gas flow to the reaction system; the magnetic drive stirring device is equipped with a polytetrafluoroethylene stirring paddle, which has strong chemical inertness and will not react with the reaction system, and the stirring rate can be adjusted steplessly in the range of 100-500 r / min to adapt to the requirements of solution mixing uniformity at different reaction stages, so as to ensure that the concentration, temperature and other parameters of the reaction system in different regions are consistent.

[0005] The laser light source is selected as a 532 nm Nd:YAG solid-state laser. The laser light with this wavelength has a high photon energy, which can excite the molecular vibration modes involved in most photocatalytic reactions. At the same time, it avoids the problems of photocatalyst photodegradation caused by shorter wavelength laser and weak Raman scattering signal caused by longer wavelength laser. The laser power is continuously adjustable in the range of 10-500 mW through an electric control module. It can be flexibly adjusted according to the response characteristics of different photocatalysts to light intensity and the needs of reaction monitoring. For example, for CdS catalyst which is sensitive to strong light and prone to photo-etching, the laser power can be adjusted to 10-50 mW. The laser spot diameter is controlled in the range of 50-200 μm through a focusing lens group. Smaller spot diameter helps to improve spatial resolution and accurately monitor the reaction changes in specific areas of the catalyst surface.

[0006] The spectrometer adopts a backscattering optical design, which can maximize the collection of Raman scattering light generated on the surface of the catalyst and improve signal strength. The resolution of the spectrometer is not less than 1 cm -1 , which can clearly distinguish the difference between adjacent characteristic peaks, such as the characteristic peaks of 144 cm -1 for TiO2 anatase phase and 133 cm -1 for rutile phase; the detection range covers 200-4000 cm -1 , which can cover the key spectral regions of catalyst lattice vibration (200-1000 cm -1 ), O-H bond vibration (3000-3600 cm -1 ) and other key spectral regions commonly used in photocatalytic hydrogen production reactions. The equipped refrigeration type CCD detector (refrigeration temperature ≤-70℃) can significantly reduce the dark current noise of the detector and improve the signal-to-noise ratio of the spectral signal, ensuring that the weak reaction intermediate signal can be detected.

[0007] The data acquisition device is equipped with a special spectral analysis software, and the sampling frequency is set to 0.1-10 Hz. When the reaction is slow, the sampling frequency of 0.1-1 Hz can be used to reduce the data volume; when the reaction is intense and changes rapidly, the sampling frequency can be increased to 5-10 Hz to avoid missing key reaction information.

[0008] During the reaction, the photocatalyst and the reaction solution (such as a mixed solution of deionized water and methanol, and methanol as a hole trapping agent) are placed in the reaction tank in a certain proportion. The gas inlet and outlet of the reaction tank are connected with a glass rotor flowmeter and a precision needle valve. The glass rotor flowmeter is used to display the gas flow in real time, and the precision needle valve is used to accurately adjust the flow, with a control accuracy of ±1 mL / min. The gas introduced is Ar, N2 inert gas with a purity of ≥99.999% and a mixture of Ar, N2 inert gas and H2. The high-purity gas can avoid the interference of impurity gases (such as CO2 and O2) on the reaction. Ar and N2 inert gas are used to remove air in the reaction system to prevent O2 from combining with photo-generated electrons. The mixed gas is used to simulate the reaction environment under different hydrogen partial pressures. The gas is treated by a dehydration and purification device (dew point ≤-40°C) before entering the reaction tank to completely remove the water in the gas and avoid the interference of water on the O-H bond characteristic peak in the Raman spectrum. The gas outlet is connected with a gas chromatograph (equipped with a TCD detector) to monitor the hydrogen production in real time, so as to synchronize the analysis of hydrogen production rate and Raman spectrum change.

[0009] Through real-time monitoring of the photocatalytic hydrogen production reaction process by the system, rich Raman spectrum data can be obtained, including: the lattice stress change caused by CdS photo-corrosion corresponding to the 400-500cm -1 characteristic peak shift (detection accuracy ±0.5cm -1 ), because the lattice stress change of CdS will cause the characteristic vibration frequency to shift, and the photo-corrosion phenomenon can be detected early by monitoring the shift; the 144cm -1 (anatase) and 133cm -1 (rutile) characteristic peak intensity change (intensity error ≤3%) related to the phase transition of TiO2, according to the relative change of the characteristic peak intensity, the content ratio of the two crystal phases can be quantitatively analyzed with the evolution of the reaction time; the characteristic peak of ·OH intermediate at 3000-3600cm -1 and the O-H bond vibration frequency shift (resolution ±0.2cm -1 ), the ·OH intermediate is a key intermediate in the photocatalytic water splitting reaction, and the appearance and intensity change of its characteristic peak directly reflect the progress of the water splitting reaction.

[0010] The special software equipped with the data acquisition device has powerful real-time drawing function, which can dynamically display the change curve of the Raman characteristic peak intensity with the reaction time (time resolution ±0.1s), which is convenient for researchers to intuitively observe the dynamic change trend of the reaction. The software also has a built-in characteristic peak drift recognition algorithm, which can automatically recognize the drift trend of the characteristic peak. When the shift of the 430cm -1 characteristic peak of CdS is detected to be more than 5cm -1When the light corrosion is detected, the light corrosion warning is triggered immediately, and the warning signal is transmitted to the electric control module of the laser light source through the software interface to control the light source power to automatically adjust (adjustment response time ≤ 1s), such as reducing the power from 300mW to 100mW, so as to effectively inhibit the light corrosion process of CdS.

[0011] S2. Constructing a molecular dynamics simulation model Based on the Raman spectrum data obtained in step S1, any one or more of the catalyst crystal phase composition, lattice parameters and bond length and bond angle information of the adsorbed molecules is extracted by peak fitting and fitting through the PeakFit software, and a molecular dynamics initial model containing the surface layer of the photocatalyst and the layer of adsorbed water molecules is constructed by using the Materials Studio software Specifically, based on the Raman spectrum data obtained in step S1, first, the PeakFit software is used to perform peak fitting and processing on the original Raman spectrum. The software uses a Gaussian-Lorentz mixed function to analyze the overlapping characteristic peaks, which can accurately extract the catalyst crystal phase composition (such as the proportion of anatase TiO2, with a calculation error ≤ 2%), the principle of which is to determine the relative content of each crystal phase by comparing the integral area of different crystal phase characteristic peaks; the lattice parameters (such as the a-axis and c-axis lattice constants of CdS, with an accuracy of ± 0.001 Å) are accurately obtained according to the correlation formula of the characteristic peak vibration frequency and the lattice constant; and the bond length and bond angle information of the adsorbed molecules is extracted, which is derived according to the theoretical relationship between the molecular vibration frequency and the bond length and bond angle.

[0012] The initial model for molecular dynamics was constructed using Materials Studio software, which included the surface layer of the photocatalyst and the adsorbed water molecule layer. The thickness of the surface layer of the photocatalyst was set to 2-5 nm, containing 3-5 atomic layers. This thickness could ensure that the simulation system covered the active sites on the catalyst surface, and also avoided the dramatic increase in computational load due to the large size of the system. The number of adsorbed water molecules was 50-200, calculated according to the actual water coverage of the reaction system, to accurately reflect the adsorption state of water molecules on the catalyst surface. The x, y, and z direction size ratio of the model was set to 1:1:(2-3), and the z direction size was larger than the x and y directions, in order to avoid the interaction between the upper and lower surfaces and simulate the reaction environment of an infinite catalyst surface. The boundary condition was a three-dimensional periodic boundary condition, which made the model have periodicity in the x, y, and z directions, and could simulate the properties of the macroscopic system. The temperature of the simulation system was set to the actual reaction temperature ±2K by the Nose-Hoover heat bath control method, which could effectively maintain the stability of the system temperature. The pressure was controlled to be 1±0.05 atm by the Parrinello-Rahman method to simulate the normal pressure reaction environment in the reaction cell, ensuring the consistency of the simulation conditions with the experimental conditions.

[0013] S3. Perform molecular dynamics simulation The ab initio molecular dynamics (AIMD) simulation method was used, which is based on quantum mechanics and can accurately describe the electronic interaction between atoms, thus accurately predicting the energy change and atomic motion trajectory of the reaction.

[0014] Specifically, the VASP 6.3.0 software package was used for simulation calculation, which has high precision and efficiency in handling solid materials and surface reaction simulation. The exchange correlation functional was selected as PBE+D3 (considering van der Waals force correction), and the PBE functional can well describe the electronic structure of most chemical reactions, while the D3 correction term is used to accurately calculate the intermolecular van der Waals force, which is crucial for describing the adsorption behavior of water molecules on the catalyst surface. The cutoff energy was set to 400-500 eV, which could ensure that the valence electron wave function was fully described, while considering the computational efficiency; the Γ point grid was used for K point sampling, which could greatly reduce the computational load while ensuring the calculation accuracy for surface model simulation.

[0015] The interatomic interactions are described by the projector augmented wave (PAW) method, which can efficiently handle the interaction between core and valence electrons and improve the computational efficiency. The simulation time step is set to 1-2 fs, which is smaller than the characteristic period of atomic vibration and can accurately capture the details of atomic motion; the total simulation time is not less than 100 ps, including a 20 ps equilibrium stage and an 80 ps production stage. The equilibrium stage is used to make the system reach a stable thermodynamic state, and the production stage is used to collect relevant data of the reaction process.

[0016] During the simulation, the bond angle changes (accuracy ±0.1°) of the H2O adsorbed around the IrRu single-atom site during the photocatalytic hydrogen production reaction process are analyzed in detail. The change of the bond angle reflects the change of the orientation of the water molecule on the catalyst surface, which is closely related to the activation process of water; the M-H bond length evolution (resolution ±0.01 Å) is tracked, and the change of the M-H bond length directly reflects the strength of the interaction between the metal site and the hydrogen atom, which is an important basis for judging the hydrogen adsorption and desorption process; the differential charge density distribution (spatial grid step ≤0.05 Å) is calculated, and the positive and negative distribution of the differential charge density can clearly determine the direction and amount of electron transfer, revealing the charge transfer mechanism in the catalytic reaction; the reaction activation energy barrier (calculated by the NEB method, accuracy ±0.01 eV) is determined, which is a key parameter to measure the difficulty of the reaction. The lower the activation energy barrier, the easier the reaction.

[0017] At the same time, the simulation also includes the calculation of the separation efficiency of photo-generated electron-hole pairs, using the Bader charge analysis method combined with the electron localization function (ELF) analysis. The Bader charge analysis method can accurately calculate the charge of an atom and determine the amount of charge transfer (accuracy ±0.01 e - ); the electron localization function (ELF) is used to describe the localization degree of electrons and assist in analyzing the separation state of electron-hole pairs. The carrier lifetime is determined by calculating the exciton recombination time (error ≤1 ps), and the longer the carrier lifetime, the more time the photo-generated electrons and holes have to participate in surface chemical reactions. Time-resolved fluorescence spectrum data are used to calibrate the carrier dynamics parameters during the simulation, making the simulation results closer to the actual experimental situation.

[0018] S4. Coupling analysis data The Raman spectrum data of S1 and the molecular dynamics simulation results of S3 are associated by using the principal component analysis (PCA) dimension reduction combined with the partial least squares regression (PLSR) data fusion algorithm, and a three-dimensional response surface model is constructed by Origin software.

[0019] Specifically, the Raman spectrum data obtained in step S1 and the molecular dynamics simulation results obtained in step S3 are deeply coupled and analyzed, which is a key link to realize precise regulation. First, principal component analysis (PCA) is used to reduce the dimension of the Raman spectrum data. PCA can extract the most representative principal components from a large amount of spectrum data, reduce data redundancy, and highlight the main features. Then, a data fusion algorithm based on partial least squares regression (PLSR) is used. PLSR can establish a nonlinear correlation model between the Raman spectrum data and the molecular dynamics simulation results, effectively dealing with the multicollinearity problem between variables.

[0020] The Raman characteristic peak intensity (error ≤5%) is associated with the simulated atomic motion trajectory (spatial resolution ±0.05 Å) and potential energy surface change (accuracy ±0.001 eV). A three-dimensional response surface model is constructed by Origin software, which can intuitively show the relationship between the characteristic peak intensity, atomic motion parameters and reaction energy. Based on this model, the generation rate (unit: mol·L -1 ·s -1 , calculation error ≤5%) of the reaction intermediates (·OH, H2O2) is accurately determined, and the size of the generation rate reflects the generation efficiency of the intermediates; the time node of bond breaking / forming is determined (accuracy ±1 ps), which helps to understand the step-by-step process of the reaction; the activation energy (error ≤0.02 eV) of the rate-determining step of the reaction is calculated, and the rate-determining step is the slowest step in the entire reaction process, and the size of its activation energy determines the rate of the entire reaction.

[0021] S5. Dynamic regulation of reaction According to the coupling analysis results, the photocatalytic hydrogen production interface reaction is dynamically regulated from multiple dimensions to improve the hydrogen production efficiency.

[0022] Specifically: ① Regulate the composition of the photocatalyst: introduce IrRu double-odd single atom sites by impregnation-reduction method. The specific process is to disperse the photocatalyst carrier (such as TiO2) in a mixed solution containing IrCl3 and RuCl3, impregnate for 12-24 h to make the metal ions fully adsorb on the surface of the carrier, then add NaBH4 reduction solution (concentration of 0.1-0.5 mol / L), and reduce at room temperature for 2-4 h to obtain a photocatalyst loaded with IrRu double-odd single atoms. Its loading amount is detected by ICP-MS to ensure that it is within the range of 0.1-1 wt% (error ≤0.01 wt%), which can ensure sufficient active sites and avoid the decrease of activity caused by metal atom aggregation.

[0023] TiO2 / CdS heterojunction is constructed by hydrothermal method. Specifically, a titanium source (tetrabutyl titanate) and a cadmium source (cadmium nitrate) are dissolved in an ethylene glycol-water mixed solvent (volume ratio 1:1) according to a Ti:Cd molar ratio of 1:(0.1-0.5). The addition of ethylene glycol can adjust the viscosity of the solution, which is conducive to the uniform mixing of reactants. 0.1-0.5mol / L thiourea is added as a sulfur source. Thiourea will slowly decompose under hydrothermal conditions to provide S 2- The mixed solution is transferred to a polytetrafluoroethylene-lined reaction kettle (filling degree 70-80%). The control of filling degree can avoid excessive pressure during the reaction. The reaction is carried out at 120-200℃ for 6-24h. Different reaction temperatures and times will affect the crystallinity and morphology of the heterojunction. After the reaction is completed, it is naturally cooled to room temperature. The product is collected by centrifugation and washed with deionized water and ethanol alternately for 3-5 times to completely remove the surface residual impurities. The product is dried in a vacuum drying oven at 60-80℃ for 12h. The vacuum environment can prevent the product from being oxidized. The high-resolution TEM characterization ensures that the contact area of the heterojunction interface is not less than 50m 2 / g (calculated by BET method combined with TEM characterization), and the larger contact area is conducive to the transfer of photo-generated charges. At the same time, the interface lattice matching degree is ensured to be ≥80% (the deviation between adjacent crystal planes is ≤10%), and good lattice matching can reduce the potential barrier of interface charge transfer.

[0024] ②Regulating the interlayer spacing of layered structure photocatalyst (MoS2): Li + ions are inserted into the interlayer of MoS2 by dispersing MoS2 powder in a hexane solution of n-butyllithium (concentration of 1.6mol / L) under argon protection and stirring for 24-48h. After the reaction is completed, the product is separated by centrifugation and washed with hexane several times to remove residual n-butyllithium. The intercalated product is dispersed in deionized water. Li + ions react with water to generate LiOH and H2, and at the same time, the interlayer spacing of MoS2 is expanded. The XRD small-angle scattering characterization precisely adjusts the interlayer spacing from the original 0.62nm to 0.7-1.0nm (accuracy ±0.01nm). The appropriate interlayer spacing is conducive to the diffusion and adsorption of water molecules.

[0025] ③Adjusting the reaction atmosphere: the H2 partial pressure is accurately controlled to be 0.1-0.5atm by a mass flow controller. Lower H2 partial pressure can reduce the reverse reaction rate of the reaction. The O2 partial pressure is controlled to be ≤0.05atm (pressure sensor accuracy ±0.001atm), which reduces the capture of photo-generated electrons by O2.

[0026] (4) Adjusting the reaction conditions: the reaction temperature is adjusted to 25-80℃ (temperature control accuracy ±0.1℃) by a constant temperature water bath device, and the reaction rate can be accelerated by appropriately increasing the temperature, but too high temperature will cause the catalyst to be deactivated; the light intensity is adjusted to 100-1000mW / cm 2 (accuracy of optical power meter detection ±5mW / cm 2 ) by adjusting the power of the laser light source or using a filter, so that the photocatalyst can fully absorb photons to generate photo-generated charges.

[0027] Through the above-mentioned multi-aspect synergistic regulation, the dynamic optimization of the photocatalytic hydrogen production interface reaction is realized, and finally the hydrogen production efficiency is improved by more than 30% (calculated in terms of hydrogen production rate, relative error ≤5%).

[0028] The above technical solutions can bring the following technical effects: The photocatalytic hydrogen production efficiency is significantly improved: through the dynamic regulation realized by in-situ Raman spectroscopy and molecular dynamics coupling simulation, the photocatalytic hydrogen production efficiency can be improved by more than 30% (calculated in terms of hydrogen production rate). Taking TiO2 / CdS heterojunction photocatalyst as an example, after optimization, its hydrogen production rate can reach 150-200μmol・g -1 ・h -1 , which is 50%-100% higher than that of the unregulated system (80-100μmol・g -1 ・h -1 ). This is due to the synergistic effect of multiple parameters: the introduction of IrRu double-odd atomic sites reduces the H2O dissociation activation energy barrier from 1.2eV to 0.8eV, the water molecule diffusion coefficient is increased by 40% after the layer spacing of layered MoS2 is regulated to 0.8nm, and the reverse reaction rate is reduced by 30% when the H2 partial pressure in the reaction atmosphere is controlled at 0.3atm. At the same time, the CdS photocorrosion rate is reduced by 60-70% after real-time monitoring and dynamic adjustment of laser power, and the service life of the catalyst is extended to 2-3 times of the original.

[0029] The interface reaction mechanism analysis accuracy is greatly improved: the coupling analysis method realizes the quantitative correlation of macroscopic spectral data and microscopic simulation results, and the reaction intermediate identification error is ≤5%, and the reaction decision step deviation is ≤0.05eV. For example, by establishing a linear relationship (R 2 =0.92) between the characteristic peak area of 3400cm -1 ·OH and the H2O dissociation rate calculated by molecular dynamics simulation, the ·OH generation rate can be accurately calculated (error ≤0.01mol・L -1 ・s -1 ); by means of the linear correlation (R 2=0.98), enabling real-time tracking of the crystal phase transformation process, with results deviating from XRD-refined data by ≤3%. Furthermore, Bader charge analysis combined with Raman characteristic peak shifting can quantitatively characterize charge transfer (accuracy ±0.01e). - This provides direct evidence for the study of photogenerated carrier separation mechanisms.

[0030] The scientific rigor and universality of the control strategy are enhanced: This method establishes a closed-loop optimization system of "spectral characteristics - microscopic mechanisms - control parameters," providing a clear theoretical basis for control measures rather than empirical adjustments. Validation on different systems demonstrates its significant universality: In the CdS system, by monitoring 430 cm⁻¹... -1 Peak shifting achieves photocorrosion suppression; in the TiO2 system, preparation conditions are optimized based on the spectral characteristics of crystal phase transformation; in the MoS2 system, interlayer vibration peaks (400-500 cm⁻¹) are utilized. -1 The interlayer spacing is adjusted by changing the parameters. Simultaneously, the quantitative standards for these parameters (e.g., heterojunction interface contact area ≥ 50m²) are also established. 2 / g, lattice matching degree ≥80%) improves the reproducibility of results from different laboratories to over 90%, providing a standardized technical path for the directional design of photocatalytic materials. Attached Figure Description

[0031] Figure 1 This is a flowchart of the dynamic control method for interfacial reactions in photocatalytic hydrogen production according to the present invention. Detailed Implementation

[0032] The following will refer to the appendices in the embodiments of the present invention. Figure 1 The technical solutions in the embodiments of the present invention are clearly and completely described herein. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0033] The embodiments provide a method for dynamically controlling the interfacial reaction in photocatalytic hydrogen production based on in-situ Raman spectroscopy-molecular dynamics coupling simulation. This method is implemented through the following detailed steps, each closely linked to form a complete dynamic control system: S1. Construct an in-situ Raman spectroscopy monitoring system (The in-situ Raman spectroscopy monitoring system is the core device for real-time monitoring. The selection and design of each component have been rigorously considered to meet the special requirements of monitoring photocatalytic hydrogen production reaction). The system comprises a corrosion-resistant light-transmitting reaction cell with a volume of 50-200 mL, and high-purity quartz glass (purity ≥ 99.9%) is selected as the material of the reaction cell. The high-purity quartz glass has excellent chemical stability and can resist corrosion of the reaction solution (such as acidic or alkaline aqueous solution). Moreover, the light transmittance of the high-purity quartz glass is as high as 90% or more in the Raman detection range of 200-4000 cm -1 , which can effectively reduce the absorption and interference of Raman scattered light. The reaction cell body is provided with a gas inlet and a gas outlet with an inner diameter of 3-5 mm. This size design can ensure smooth gas flow and reduce the disturbance of gas flow to the reaction system. The magnetic drive stirring device is equipped with a polytetrafluoroethylene stirring paddle. This material has strong chemical inertness and will not chemically react with the reaction system. The stirring rate can be adjusted steplessly in the range of 100-500 r / min to meet the requirements of different reaction stages for solution mixing uniformity, and to ensure that the concentration, temperature and other parameters of each region of the reaction system remain consistent.

[0034] The laser light source is an Nd:YAG solid-state laser with a wavelength of 532 nm. This wavelength of laser has high photon energy and can excite most of the molecular vibration modes involved in photocatalytic reactions. At the same time, it avoids the problems of photocatalyst light degradation caused by shorter wavelength laser and weak Raman scattering signal caused by longer wavelength laser. The laser power is continuously adjustable in the range of 10-500 mW through an electric control module. It can be flexibly adjusted according to the response characteristics of different photocatalysts to light intensity and the needs of reaction monitoring. For example, for CdS catalyst which is sensitive to strong light and prone to photo-corrosion, the laser power can be adjusted to 10-50 mW. The laser spot diameter is controlled to be 50-200 μm through a focusing lens group. Smaller spot diameter helps to improve the spatial resolution and accurately monitor the reaction changes in specific areas of the catalyst surface.

[0035] The spectrometer adopts a backscattering light path design, which can maximize the collection of Raman scattered light generated on the surface of the catalyst and improve the signal strength. The resolution of the spectrometer is not less than 1 cm -1 , which can clearly distinguish the difference between adjacent characteristic peaks, such as the characteristic peaks of TiO2 anatase phase 144 cm -1 and rutile phase 133 cm -1 . The detection range covers 200-4000 cm -1 , which can cover the common catalyst lattice vibrations (200-1000 cm -1 ), O-H bond vibrations (3000-3600 cm -1) and other key spectral regions. Equipped with a refrigerated CCD detector (refrigeration temperature ≤-70℃), the dark current noise of the detector can be significantly reduced, and the signal-to-noise ratio of the spectral signal can be improved to ensure that the weak reaction intermediate signal can be detected.

[0036] The data acquisition device is equipped with special spectral analysis software, and the sampling frequency is set to 0.1-10Hz. When the reaction is slow, a sampling frequency of 0.1-1Hz can be used to reduce the amount of data; when the reaction is intense and changes rapidly, the sampling frequency can be increased to 5-10Hz to avoid missing key reaction information.

[0037] During the reaction, the photocatalyst and the reaction solution (such as a mixed solution of deionized water and methanol, with methanol as a hole trapping agent) are placed in the reaction cell in a certain proportion. The gas inlet and outlet of the reaction cell are connected in series with a glass rotor flowmeter and a precision needle valve. The glass rotor flowmeter is used to display the gas flow in real time, and the precision needle valve is used to accurately adjust the flow, with a control accuracy of ±1mL / min. The gas introduced is Ar, N2 inert gas with a purity of ≥99.999% and a mixture of Ar, N2 inert gas and H2. High-purity gas can avoid the interference of impurity gases (such as CO2, O2) on the reaction. Ar, N2 inert gas is used to exclude air in the reaction system to prevent O2 from reacting with photo-generated electrons. The mixed gas is used to simulate the reaction environment under different hydrogen partial pressures. The gas is treated by a dehydration purification device (dew point ≤-40℃) before entering the reaction cell to completely remove the water in the gas and avoid the interference of water on the O-H bond characteristic peak in the Raman spectrum. The gas outlet is connected to a gas chromatograph (equipped with a TCD detector) to monitor the hydrogen production in real time, so that the hydrogen production rate and the Raman spectrum change can be analyzed synchronously.

[0038] Through real-time monitoring of the photocatalytic hydrogen production reaction process by the system, a wealth of Raman spectral data can be obtained, including: the lattice stress change caused by CdS photo-corrosion corresponds to the 400-500cm -1 peak shift (detection accuracy ±0.5cm -1 ), which is because the lattice stress change of CdS will cause the characteristic vibration frequency to shift, and by monitoring the shift, the photo-corrosion phenomenon can be detected early; the intensity change of the 144cm -1 (anatase) and 133cm -1 (rutile) characteristic peaks (intensity error ≤3%) related to TiO2 crystal phase transition, according to the relative change of the characteristic peak intensity, the content ratio of the two crystal phases can be quantitatively analyzed with the evolution of the reaction time; the characteristic peak of ·OH intermediate at 3000-3600cm -1 and the O-H bond vibration frequency shift (resolution ±0.2cm -1OH intermediate is a key intermediate in the photocatalytic water splitting reaction, and the appearance and intensity change of its characteristic peak directly reflect the progress of the water splitting reaction.

[0039] At the same time, the special software equipped with the data acquisition device has powerful real-time drawing function, which can dynamically display the change curve of Raman characteristic peak intensity with reaction time (time resolution ±0.1s), and facilitate researchers to observe the dynamic change trend of the reaction intuitively. The software also has a built-in characteristic peak drift recognition algorithm, which can automatically identify the drift trend of the characteristic peak. When the drift of the characteristic peak of CdS at 430cm -1 characteristic peak drift exceeds 5cm -1 , the photo-etching warning is triggered immediately, and the warning signal is transmitted to the electric control module of the laser light source through the software interface to control the automatic adjustment of the light source power (adjustment response time ≤1s), such as reducing the power from 300mW to 100mW, so as to effectively inhibit the photo-etching process of CdS.

[0040] S2. Constructing molecular dynamics simulation model Based on the Raman spectrum data obtained in step S1, first, the PeakFit software is used to perform peak fitting processing on the original Raman spectrum. The software uses Gaussian-Lorentz mixed function to analyze the overlapping characteristic peaks, which can accurately extract the catalyst crystal phase composition (such as the proportion of anatase TiO2, the calculation error is ≤2%), the principle is to determine the relative content of each crystal phase by comparing the integral area of different crystal phase characteristic peaks; accurately obtain the lattice parameters (such as the a-axis and c-axis lattice constants of CdS, accuracy ±0.001Å), which are calculated according to the correlation formula of characteristic peak vibration frequency and lattice constant; at the same time, the bond length and bond angle information of the adsorbed molecule is extracted, which is derived according to the theoretical relationship between molecular vibration frequency and bond length and bond angle.

[0041] The initial model for molecular dynamics was constructed using Materials Studio software, which included the surface layer of the photocatalyst and the adsorbed water molecule layer. The thickness of the surface layer of the photocatalyst was set to 2-5 nm, containing 3-5 atomic layers. This thickness could ensure that the simulation system covered the active sites on the catalyst surface, and also avoided the dramatic increase in computational load due to the large size of the system. The number of adsorbed water molecules was 50-200, calculated according to the actual water coverage of the reaction system, to accurately reflect the adsorption state of water molecules on the catalyst surface. The x, y, and z direction size ratio of the model was set to 1:1:(2-3), and the z direction size was larger than the x and y directions, in order to avoid the interaction between the upper and lower surfaces and simulate the reaction environment of an infinite catalyst surface. The boundary condition was three-dimensional periodic boundary condition, which made the model have periodicity in x, y, and z directions, and could simulate the properties of the macroscopic system. The temperature of the simulation system was set to the actual reaction temperature ±2K by the Nose-Hoover heat bath control method, which could effectively maintain the stability of the system temperature. The pressure was controlled to 1±0.05 atm by the Parrinello-Rahman method to simulate the normal pressure reaction environment in the reaction cell, ensuring the consistency of the simulation conditions with the experimental conditions.

[0042] S3. Perform molecular dynamics simulation The ab initio molecular dynamics (AIMD) simulation method was used, which was based on quantum mechanics and could accurately describe the electronic interaction between atoms, thus accurately predicting the energy change and atomic motion trajectory of the reaction. The VASP6.3.0 software package was used for simulation calculation, which had high precision and efficiency in handling solid materials and surface reaction simulation. The exchange correlation functional was PBE+D3 (considering van der Waals force correction), which could well describe the electronic structure of most chemical reactions, and the D3 correction term was used to accurately calculate the intermolecular van der Waals force, which was crucial for describing the adsorption behavior of water molecules on the catalyst surface. The cutoff energy was set to 400-500 eV, which could ensure that the valence electron wave function was fully described, while considering the calculation efficiency; the Γ point grid was used for K point sampling, which could greatly reduce the computational load while ensuring the calculation accuracy for surface model simulation.

[0043] The interatomic interactions are described by the projector augmented wave (PAW) method, which can efficiently handle the interaction between core and valence electrons and improve the computational efficiency. The simulation time step is set to 1-2 fs, which is smaller than the characteristic period of atomic vibration and can accurately capture the details of atomic motion; the total simulation time is not less than 100 ps, including a 20 ps equilibrium stage and an 80 ps production stage. The equilibrium stage is used to make the system reach a stable thermodynamic state, and the production stage is used to collect relevant data of the reaction process.

[0044] During the simulation, the bond angle changes (accuracy ±0.1°) of the H2O adsorbed around the IrRu single-atom site during the photocatalytic hydrogen production reaction process are analyzed in detail. The change of the bond angle reflects the change of the orientation of the water molecule on the catalyst surface, which is closely related to the activation process of water; the M-H bond length evolution (resolution ±0.01 Å) is tracked, and the change of the M-H bond length directly reflects the strength of the interaction between the metal site and the hydrogen atom, which is an important basis for judging the hydrogen adsorption and desorption process; the differential charge density distribution (spatial grid step ≤0.05 Å) is calculated, and the positive and negative distribution of the differential charge density can clearly determine the direction and amount of electron transfer, revealing the charge transfer mechanism in the catalytic reaction; the reaction activation energy barrier (calculated by the NEB method, accuracy ±0.01 eV) is determined, which is a key parameter to measure the difficulty of the reaction. The lower the activation energy barrier, the easier the reaction.

[0045] At the same time, the simulation also includes the calculation of the separation efficiency of photo-generated electron-hole pairs, using the Bader charge analysis method combined with the electron localization function (ELF) analysis. The Bader charge analysis method can accurately calculate the charge of the atom and determine the amount of charge transfer (accuracy ±0.01 e - ); the electron localization function (ELF) is used to describe the localization degree of electrons, which helps to analyze the separation state of electron-hole pairs. The carrier lifetime is determined by calculating the exciton recombination time (error ≤1 ps), and the longer the carrier lifetime, the more time the photo-generated electrons and holes have to participate in surface chemical reactions. Time-resolved fluorescence spectrum data are used to calibrate the carrier dynamics parameters during the simulation, making the simulation results closer to the actual experimental situation.

[0046] S4. Coupling analysis data The Raman spectrum data obtained in step S1 and the molecular dynamics simulation results obtained in step S3 are deeply coupled and analyzed, which is a key link to realize precise regulation. First, principal component analysis (PCA) is used to reduce the dimension of the Raman spectrum data. PCA can extract the most representative principal components from a large amount of spectrum data, reduce data redundancy, and highlight the main features. Then, a data fusion algorithm based on partial least squares regression (PLSR) is used. PLSR can establish a nonlinear correlation model between the Raman spectrum data and the molecular dynamics simulation results, effectively dealing with the multicollinearity problem between variables.

[0047] The Raman characteristic peak intensity (error ≤5%) is associated with the simulated atomic motion trajectory (spatial resolution ±0.05Å) and potential energy surface change (accuracy ±0.001eV). A three-dimensional response surface model is constructed by Origin software, which can intuitively show the relationship between the characteristic peak intensity, atomic motion parameters and reaction energy. Based on this model, the generation rate (unit: mol・L -1 ・s -1 , calculation error ≤5%) of the reaction intermediates (·OH, H2O2) is accurately determined, which reflects the generation efficiency of the intermediates; the time node of bond breaking / formation (accuracy ±1ps) is determined, which helps to understand the step-by-step process of the reaction; the activation energy (error ≤0.02eV) of the rate-determining step of the reaction is calculated, which determines the rate of the entire reaction.

[0048] S5. Dynamic regulation of reaction According to the coupling analysis results, the photocatalytic hydrogen production interface reaction is dynamically regulated from multiple dimensions to improve the hydrogen production efficiency: ① Regulate the composition of the photocatalyst: introduce IrRu double-odd single atom sites by impregnation-reduction method. Specifically, disperse the photocatalyst carrier (such as TiO2) in a mixed solution containing IrCl3 and RuCl3, impregnate for 12-24h to make the metal ions fully adsorb on the carrier surface, then add NaBH4 reduction solution (concentration of 0.1-0.5mol / L), reduce at room temperature for 2-4h to obtain a photocatalyst loaded with IrRu double-odd single atoms. Adjust the loading amount by ICP-MS detection combined with the concentration of metal precursor solution and the mass ratio of carrier to ensure that the loading amount is within 0.1-1wt% (error ≤0.01wt%), which can ensure enough active sites and avoid the reduction of activity caused by metal atom aggregation.

[0049] TiO2 / CdS heterojunction is constructed by hydrothermal method. Specifically, a titanium source (tetrabutyl titanate) and a cadmium source (cadmium nitrate) are dissolved in an ethylene glycol-water mixed solvent (volume ratio 1:1) according to a Ti:Cd molar ratio of 1:(0.1-0.5). The addition of ethylene glycol can adjust the viscosity of the solution, which is conducive to the uniform mixing of reactants. 0.1-0.5 mol / L thiourea is added as a sulfur source. Thiourea will slowly decompose under hydrothermal conditions to provide S 2- The mixed solution is transferred to a polytetrafluoroethylene-lined reaction kettle (filling degree 70-80%). The control of filling degree can avoid excessive pressure during the reaction. The reaction is carried out at 120-200 ℃ for 6-24 h. Different reaction temperatures and times will affect the crystallinity and morphology of the heterojunction. After the reaction is completed, it is naturally cooled to room temperature. The product is collected by centrifugation and washed with deionized water and ethanol alternately for 3-5 times to completely remove the surface residual impurities. The product is dried in a vacuum drying oven at 60-80 ℃ for 12 h. The vacuum environment can prevent the product from being oxidized. The high-resolution TEM characterization ensures that the interface contact area of the heterojunction is not less than 50 m 2 / g (calculated by BET method combined with TEM characterization), and a larger contact area is conducive to the transfer of photo-generated charges. At the same time, it is ensured that the interface lattice matching degree is ≥80% (the deviation of the distance between adjacent crystal planes is ≤10%). Good lattice matching can reduce the potential barrier of interface charge transfer. By systematically optimizing the hydrothermal synthesis parameters, the interface contact area and lattice matching degree of the heterojunction can be controlled. For example, experiments show that when the molar ratio of titanium source to cadmium source is 1:0.3, the thiourea concentration is 0.3 mol / L, the reaction temperature is 180 ℃, and the reaction time is 12 h, the BET specific surface area of the obtained TiO2 / CdS heterojunction is 120 m 2 / g, the estimated interface contact area is 65 m 2 / g, and the high-resolution TEM shows that the distance deviation between TiO2(101) and CdS(002) crystal planes is 6%, and the lattice matching degree is 88%, meeting the requirements of contact area ≥50 m² / g and matching degree ≥80%. If the contact area is low, it can be further improved by extending the reaction time to 18 h or adding 0.01 mol / L citric acid as an interface coupling agent.

[0050] ②Regulating the interlayer distance of layered structure photocatalyst (MoS2): MoS2 powder is dispersed in n-butyllithium hexane solution (concentration of 1.6 mol / L) by Li⁺ ion intercalation method. Under argon protection, stir for 24-48 h. Li⁺ ions will intercalate into the interlayer of MoS2. After the reaction is completed, the product is separated by centrifugation and washed with hexane several times to remove residual n-butyllithium. The intercalated product is dispersed in deionized water. Li +The ions react with water to generate LiOH and H2, and at the same time, the interlayer spacing of MoS2 is expanded. The interlayer spacing is precisely adjusted from the original 0.62 nm to 0.7-1.0 nm (accuracy ±0.01 nm) through XRD small-angle scattering characterization. The appropriate interlayer spacing is conducive to the diffusion and adsorption of water molecules.

[0051] Specifically, the n-butyllithium intercalation-water exfoliation method is used to realize the controllable adjustment of the interlayer spacing of MoS2: (1) 0.5 g of MoS2 powder is dispersed in 50 mL of 1.6 mol / L n-butyllithium-n-hexane solution, and stirred at 25°C for 24-48 h under argon protection; (2) After centrifugal separation, the product is quickly dispersed in 100 mL of deionized water, and treated with 200 W ultrasonic for 30 min, Li + reacts with water to generate LiOH and H2, and causes the interlayer expansion; (3) The (002) peak shift is monitored in real time through XRD small-angle scattering (2θ = 3-10°), and the ultrasonic treatment is immediately terminated when the diffraction peak is shifted to 2θ≈8.8° (corresponding to d = 1.0 nm); (4) If the interlayer spacing exceeds the target value, the interlayer spacing can be shrunk by 0.02-0.05 nm through annealing at 120°C for 1 h, and finally the precise control of ±0.01 nm within the range of 0.70-1.00 nm is realized.

[0052] ③Adjusting the reaction atmosphere: the H2 partial pressure is accurately controlled to be 0.1-0.5 atm through a mass flow controller, and the lower H2 partial pressure can reduce the reverse reaction rate of the reaction; the O2 partial pressure is controlled to be ≤0.05 atm (pressure sensor accuracy ±0.001 atm), which reduces the capture of photo-generated electrons by O2.

[0053] ④Adjusting the reaction conditions: the reaction temperature is adjusted to be 25-80°C (temperature control accuracy ±0.1°C) through a constant temperature water bath device, and the appropriate increase of temperature can accelerate the reaction rate, but the excessive high temperature will cause the deactivation of the catalyst; the light intensity is adjusted to be 100-1000 mW / cm 2 (measured by a light power meter with an accuracy of ±5 mW / cm 2 ) by adjusting the power of the laser light source or using a filter, so that the photocatalyst can fully absorb photons to generate photo-generated charges.

[0054] Through the above-mentioned multi-faceted synergistic regulation, the dynamic optimization of the interface reaction of photocatalytic hydrogen production is realized, and finally the hydrogen production efficiency is improved by more than 30% (calculated by the hydrogen production rate, with a relative error of ≤5%).

[0055] More specific embodiments are as follows: Example 1: Regulation of interfacial reactions in photocatalytic hydrogen production based on CdS photocatalyst Experimental Setup: The in-situ Raman spectroscopy monitoring system was constructed strictly according to step S1 above. A 100mL high-purity quartz glass cell (99.99% purity) was used as the reaction cell. The inner diameter of both the inlet and outlet of the cell was 4mm. A three-bladed PTFE stirrer (20mm diameter) was used, and the stirring speed was set to 300r / min. A 532nm Nd:YAG laser was used as the laser source, with the initial power adjusted to 30mW (calibrated by a power meter, error ±1mW). The laser spot was compressed to 100μm by a focusing lens group (verified using a laser rangefinder). The spectrometer resolution was set to 0.8cm⁻¹. -1 The detection range covers 200-4000cm. -1 The cooled CCD detector operates at a stable temperature of -70℃, and the data acquisition software is set to a sampling frequency of 2Hz to ensure that a complete spectrum is recorded every 5 seconds.

[0056] Reaction system preparation: Weigh 0.1g of CdS photocatalyst (average particle size 30nm, particle size distribution range ≤1.2 as determined by dynamic light scattering), add 100mL of 10% methanol-water solution (methanol volume fraction calibrated by gas chromatography, error ±0.5%), and ultrasonically disperse for 30min (power 300W, frequency 40kHz) until the solution forms a homogeneous suspension. After transferring the suspension to the reaction tank, 99.999% pure Ar gas (treated by a dehydration and purification device, dew point ≤-45℃) is introduced. The gas flow rate is stabilized at 50mL / min by a mass flow controller, and continuous aeration is carried out for 30min to completely remove air from the tank (O2 residue ≤5ppm as detected by gas chromatography).

[0057] Real-time monitoring and simulation: 300W xenon lamp (equipped with AM 1.5G filter, light intensity calibrated to 100mW / cm²) is turned on. 2 The temperature of the reaction tank was controlled at 25.0 ± 0.1 ℃ using a constant-temperature circulating water bath, and the total pressure inside the tank was maintained at 1.00 ± 0.02 atm (adjusted by a back pressure valve and a mass flow controller). Temperature and pressure data were recorded in real time. The reaction was initiated by irradiation of the reaction tank, and in-situ Raman spectroscopy was used to track the 430 cm⁻¹ in real time. -1 The change in the characteristic peak of the Cd-S bond. Within the first 10 minutes of the reaction, the position of this characteristic peak stabilizes at 430.2 cm⁻¹. -1 The full width at half maximum (FWHM) was 8.5 cm⁻¹; after 30 min of reaction, the peak position began to shift to higher wavenumbers at a rate of 0.1 cm⁻¹. -1 / min. Meanwhile, a molecular dynamics model was constructed according to the S2-S3 steps: a CdS (0001) crystal surface was used as the substrate (containing 5 layers of Cd atoms and 5 layers of S atoms, with a layer thickness of 3.2 nm), and 80 water molecules were adsorbed on the surface (with a surface coverage of 1.2 molecules / nm 2 , the model had a size of 4 nm in the x and y directions and a size of 10 nm in the z direction (including a 5 nm vacuum layer). AIMD simulation was performed using the VASP software, the functional was selected as PBE+D3, the cutoff energy was 450 eV, the time step was 1 fs, and the total simulation time was 100 ps (the first 20 ps was an NVT equilibrium stage, and the last 80 ps was an NPT production stage). The simulation results showed that the Cd-S bond length was stable at 2.52 Å at the beginning of the reaction, and gradually increased to 2.55 Å at 30 ps, which was consistent with the trend of the Raman peak shift.

[0058] Coupling analysis and regulation: at 120 min of the reaction, the Raman spectrum monitored a 430 cm -1 peak shift of 6 cm -1 ( the actual peak position was 436.2 cm -1 ), and the peak intensity decreased by 15% compared with the initial value through peak fitting by the PeakFit software. Coupling analysis found that the shift corresponded to an increase in the Cd-S bond length in the simulation system to 2.58 Å, and the calculated lattice stress increased by 25% (obtained by the built-in stress calculation module of VASP). After the laser power was reduced to 15 mW within 1 s, the reaction path was globally optimized, and the following regulation was continued: ① Heterostructure construction: TiO2 / CdS heterojunction was introduced in situ according to the method of Example 4 (Ti:Cd=1:0.3, 180 ℃ hydrothermal for 12 h), the interface contact area increased to 65 m² / g, the lattice matching degree reached 88%, and the photo-induced charge separation efficiency increased by 40%; ② Reaction condition optimization: the H2 partial pressure was adjusted to 0.3 atm by a mass flow controller (to reduce the reverse reaction), and the constant temperature water bath was heated to 60 ℃ (to accelerate the surface reaction kinetics). The hydrogen production rate (detected by gas chromatography every 30 min) steadily increased from the initial 80 μmol·g -1 ·h -1 to 130 μmol·g -1 ·h -1 , and the activity retention rate of the catalyst was still 80% after 120 h of continuous reaction (the activity retention rate of the unregulated group was only 30% after 50 h).

[0059] Example 2: Photocatalytic hydrogen production under the regulation of TiO2 crystal phase transformation Experimental setup: In-situ Raman spectroscopy monitoring system was built, the volume of the reactor was 150 mL, the laser power was adjusted to 200 mW (spot diameter 150 pm), the spectrometer focused on scanning 100-200 cm -1 Interval (resolution 0.5 cm -1 ), to accurately capture the anatase TiO2 phase 144 cm -1 and the rutile phase 133 cm -1 characteristic peaks. The stirring rate was 250 r / min, and the air flow rate was 40 mL / min (99.999% N2).

[0060] Catalyst preparation and reaction: Anatase TiO2 was prepared by sol-gel method: 10 mL of tetrabutyl titanate was slowly added to a mixture of 50 mL of anhydrous ethanol and 5 mL of deionized water (stirring rate 500 r / min), and the pH was adjusted to 2.0 by adding nitric acid. After gelation, it was calcined at 500°C for 2h (heating rate 5°C / min) to obtain a powder with an average particle size of 20 nm (XRD characterization of anatase purity 98%). 0.15 g of the powder was dispersed in 150 mL of deionized water (conductivity ≤1 pS / cm), and after ultrasonic dispersion for 20 min, it was transferred to the reactor, and the reaction was started under 200 mW / cm 2 Xenon lamp illumination.

[0061] Simulation and analysis: Based on the Raman spectroscopy data, a molecular dynamics model was constructed, which included 5 layers of TiO2(101) crystal surface (thickness 4 nm) and 100 water molecules adsorbed on the surface. Taking TiO2(101) surface as the substrate, x, y direction 2.95 nm × 3.78 nm (corresponding to 6 × 4 anatase unit cells), a total of 5 layers of Ti atoms, thickness ≈ 4 nm; 2 nm vacuum layer on top. 100 water molecules were uniformly placed on the surface, arranged according to 1 ML coverage. The overall model size was 2.95 nm × 3.78 nm × 10 nm, and three-dimensional periodic boundary conditions were used. AIMD simulation showed that the anatase phase characteristic peak 144 cm -1 was stable (I144=1200 counts) in the initial stage of the reaction (0-60 min), and the rutile phase 133 cm -1Weak peak intensity (I133=500 counts), I144 / I133=2.4. After 120 min of reaction, I144 dropped to 900 counts, I133 rose to 750 counts, I144 / I133=1.2, and the simulated atomic trajectory showed that about 15% of the anatase lattice was restructured (Ti-O bond length increased from 1.96 Å to 2.01 Å). At this time, the temperature of the reaction cell was raised from room temperature (25°C) to 180°C (constant temperature bath temperature control accuracy ±0.5°C) by program control, and the hydrothermal environment was maintained for 30 min.

[0062] Regulation effect: ① Temperature: constant temperature circulating water bath, 5 ℃ min -1 rate, 180.0±0.5 ℃ in 30 s, and maintained for 30 min; then 3 ℃ min -1 decrease back to 25 ℃, complete a "thermal shock-quenching" cycle, a total of 2 cycles (total heat treatment 60 min).

[0063] ② Pressure: back pressure valve and mass flow controller linkage, maintain the total pressure of the system 1.00±0.02 atm; first pass 99.999% N230 min to exclude O2, then switch to N2 / H2=7:3 mixed gas, accurate control H2 partial pressure 0.30±0.01 atm, O2 partial pressure ≤50 ppm (online oxygen analyzer real-time monitoring).

[0064] ③ Illumination: constant xenon lamp power is 200 mW cm -1 (intensity meter point-by-point calibration, error ≤3%), to ensure that the light flux is constant, to exclude the influence of light intensity fluctuation on the rate.

[0065] ④ Sampling: hydrogen production rate every 10 min automatic sampling (GC-TCD), take the average of the last 3 times; Raman 144 cm -1 / 133 cm -1 intensity ratio is recorded synchronously, used to verify the stability of the crystal phase.

[0066] After 2 rounds of the above regulation, the hydrogen production rate gradually increased from 90 μmol·g -1 ·h -1 to 160 μmol·g -1 ·h -1 , an increase of 78%; XRD verified that the anatase retention rate was ≥85%, consistent with the Raman intensity ratio I 144 / I 133 =1.7, error ≤3%.

[0067] Example 3: Hydrogen production regulation of IrRu double-odd single atom site modified TiO2 Catalyst preparation: 0.5wt% IrRu / TiO2 was prepared. 1g TiO2 (specific surface area 80m 2 / g) was dispersed in 50mL mixed solution (containing 0.01mol / L IrCl2 and 0.01mol / L RuCl2, molar ratio 1:1, concentration labeled by ICP-MS), room temperature immersion for 18h (stirring for 10min every 2h). 0.2mol / L NaBH4 solution was added (molar ratio NaBH4:Ir+Ru=5:1), magnetic stirring for 3h (rotation speed 400r / min), centrifugal collection of solid (8000r / min, 10min), washed with deionized water for 5 times until the filtrate was free of Cl⁻ (AgNO3 detection). ICP-MS detection of actual loading was 0.495wt%, error ±0.005wt%, XPS characterization showed that Ir 0 / Ru 0 accounted for 96%. The target loading was quantitatively controlled mainly by adjusting the concentration of metal precursor solution and the mass ratio of carrier, ICP-MS detection error ≤5%, if the loading is low, the metal source can be added for secondary immersion.

[0068] Reaction monitoring: in-situ system was built, reaction cell volume 120mL, laser power 150mW, focused on the active area of catalyst surface, focusing on monitoring the characteristic peak of ·OH intermediate at 3400cm -1 (-half peak width about 60cm -1 ). The reaction solution was 150mL deionized water, Ar gas was introduced (flow rate 35mL / min), light intensity 150mW / cm 2 .

[0069] Simulation and regulation: The model of IrRu single-atom modified TiO2(101) surface was constructed by AIMD simulation, and 60 H2O molecules were adsorbed on the surface. A 4x3 TiO2(101) unit cell (x = 2.95 nm, y = 3.78 nm) was used as the substrate, with a total of 5 layers of TiO2, and the thickness was approximately 0.95 nm; a 2.5 nm vacuum layer was left on top, and the overall z was approximately 10 nm. One IrRu dimer (Ir-Ru distance 2.7 Å) was placed at the surface oxygen vacancy site, and then 60 H2O molecules were randomly placed around it, corresponding to a 0.8 ML coverage. The model size was 2.95 nm x 3.78 nm x 10 nm, and three-dimensional periodic boundary conditions were used, with a Nose-Hoover heat bath temperature control of 300 K. The simulation showed that the H2O molecule bond angle around the IrRu site gradually increased from 104.5° in the free state to 109° in the activated state, and the M-H bond length was stable at 1.6 Å (fluctuation ± 0.01 Å). The coupling analysis used the PLSR algorithm, and the 3400 cm -1 -1 peak area (recorded every 10 min, error ≤ 3%) was correlated with the simulated ·OH generation rate, resulting in a linear equation y = 0.02x + 0.5 (R 2 = 0.93). When the ·OH generation rate reached 0.08 mol·L -1 ·s -1 -1, the H2 partial pressure was adjusted from 0 to 0.3 atm by a mass flow controller, and the specific coordination parameters were as follows: ① Temperature and pressure: a constant temperature circulating water bath was used to maintain the reaction cell at 25.0 ± 0.1 ℃; a back pressure valve was connected to a mass flow controller, with a total pressure of 1.00 ± 0.02 atm and an H2 partial pressure of 0.30 ± 0.01 atm (N2 / H2 = 7:3), O2≤ 5 ppm (online oxygen analyzer); gas flow rate 40 mL min -1 -1, and data was recorded after 20 min of stable switching.

[0070] ② Illumination and stirring: a xenon lamp 300 W with an AM 1.5G filter, light intensity 150 mW cm -2 -2 (light power meter point-by-point calibration, error ≤ 3%); magnetic drive polytetrafluoroethylene stirring paddle 300 r min -1 -1, to ensure uniform suspension of the catalyst.

[0071] ③ Catalyst and solution: 0.10 g 0.495 wt % IrRu / TiO2(ICP-MS confirmed loading ±0.005 wt %) dispersed in 100 mL deionized water (conductivity ≤ 1 μS cm -1 ) after 10 min of sonication was transferred to the reaction cell; N2 was bubbled through for 30 min to remove air before reaction.

[0072] ④ Sampling and averaging: Gas chromatograph auto-sampling every 10 min, 6 stable values (≤ 2 % relative deviation within 30 min) were taken continuously to calculate the average rate; Raman 3400 cm -1 peak area was recorded synchronously for PLSR linear calibration (R 2 = 0.93).

[0073] Under this condition, the hydrogen production rate reached 190 μmol·g -1 ·h -1 , which was 123 % higher than that of pure TiO2(85 μmol·g -1 ·h -1 ); the activation energy decreased to 0.77 eV, the Bader charge transfer was 0.25 e - , and the deviation from simulation was ≤ 0.02 eV.

[0074] Example 4: TiO2 / CdS heterojunction photocatalytic hydrogen production regulation TiO2 / CdS heterojunction preparation: 5 mL tetrabutyl titanate and 0.01 mol cadmium nitrate (Ti:Cd molar ratio 1:0.3) were dissolved in 50 mL ethylene glycol-water mixture (volume ratio 1:1), 0.3 mol / L thiourea (S:Cd = 2:1) was added after stirring for 30 min, and the solution was continuously stirred for 1 h until it became a light yellow transparent solution. It was transferred to a 50 mL polytetrafluoroethylene-lined reaction kettle (filling degree 75%), and reacted at 180℃ for 12 h (oven temperature control accuracy ±2℃). After natural cooling, centrifugation (8000 r / min, 10 min), and washing with deionized water and ethanol alternately 3 times (10 mL each time), it was vacuum dried at 60℃ for 12 h (vacuum degree -0.095 MPa). HRTEM characterization showed that the contact area of the heterojunction interface was 65 m 2 / g, the matching degree of the TiO2(101) crystal plane (d = 0.35 nm) and the CdS(002) crystal plane (d = 0.33 nm) was 88% (spacing deviation 6%).

[0075] In-situ monitoring: reaction cell volume 100 mL, laser power 100 mW, sampling frequency 5 Hz, TiO2 144 cm-1 CdS 430cm -1 Characteristic peak. Reaction system: 0.1 g of heterojunction dispersed in 100 mL of 10% methanol-water, Ar flow rate 45 mL / min, light intensity 200 mW / cm 2 .

[0076] Simulation and regulation: the molecular dynamics model contains a TiO2 / CdS heterojunction interface (5 atomic layers each), with 70 H2O molecules adsorbed on the surface. The simulation shows that the amount of charge transfer at the interface reaches 0.3e - (Bader analysis), and when the Raman monitoring of CdS 430cm -1 peak shifts by 2.5cm -1 , the laser power is automatically reduced from 100mW to 80mW.

[0077] Effect: the hydrogen production rate is stabilized at 170μmol·g -1 ·h -1 , which is 95% higher than that of single TiO2 (87μmol·g -1 ·h -1 ) and 80% higher than that of single CdS (94μmol・g⁻¹・h⁻¹). Raman spectrum shows that the maximum shift of 430cm -1 peak is ≤3cm -1 , and the light corrosion rate (Cd 2+ leaching) is 0.006mg・L -1 ・h -1 , which is 70% lower than that of unregulated heterojunction.

[0078] Example 5: Layered MoS2 interlayer spacing regulated hydrogen production reaction Catalyst treatment: Li⁺ ion intercalation method was used to regulate the interlayer spacing of MoS2. 0.5g of MoS2 (original interlayer spacing 0.62nm, XRD small angle peak at 14.3°) was dispersed in 50mL of 1.6mol / L n-butyllithium-hexane solution, stirred at 25℃ for 36h (speed 300r / min) under argon protection. Centrifugal separation (5000r / min, 5min), washed with hexane 5 times (20mL each time), vacuum dried for 6h. The product was dispersed in 100mL of deionized water, ultrasonic treated for 30min (power 200W), centrifuged to collect the solid, dried at 60℃ for 12h. XRD characterization shows that the (002) peak shifts to 10.4°, and the calculated interlayer spacing is 0.85nm (accuracy ±0.005nm).

[0079] Reaction system: 80 mL reaction cell, 80 mW laser power, focused on the MoS2 surface, monitored at 450 cm. -1 Interlayer vibration peak (half-peak width 12cm) -1 The reaction system consisted of 0.08 g of modified MoS2 dispersed in 80 mL of 5% lactic acid-water solution (the lactic acid concentration was titrated and standardized), with a N2 gas flow rate of 30 mL / min and an initial temperature of 25 °C.

[0080] Simulation and Analysis: A molecular dynamics model was used to construct a layered MoS2 structure with an interlayer spacing of 0.85 nm, with 50 H2O molecules adsorbed between the layers. Simulations showed that the water molecule diffusion coefficient increased from 1.2 × 10⁻⁶ to 1.2 × 10⁻⁶ after the interlayer spacing was increased. -9 m 2 / s increased to 1.7×10⁻ 9 m² / s (increased by 42%), and 450cm² / s was improved through principal component analysis. -1 Peak intensity (recorded every 15 minutes, error ≤4%) was correlated with interlayer spacing, and the 3400 cm⁻¹ interval was used as the reference. -1 The characteristic peak area S of ·OH (in counts) and the ·OH formation rate r calculated by AIMD (in mol L) -1 s -1 Performing linear regression yields: r = 0.018 S + 0.00042, and the correlation coefficient R0. 2 =0.91, standard error σ = 0.0003 mol L -1 s -1 A linear relationship (R) is obtained. 2 =0.91). When the calculated diffusion coefficient reaches the threshold of 1.7 × 10⁻⁶. -9 m 2 s -1 At that time, the reaction temperature was raised from 25℃ to 60℃ using a constant temperature bath (heating rate 2℃ / min).

[0081] Results: After regulation, the hydrogen production rate increased from the initial 75 μmol / g. -1 h -1 Gradually increase to 130 μmolg -1 h -1 An increase of 73%. Raman spectroscopy at 450cm. -1 The peak intensity remained stable (fluctuation ≤5%), indicating that the interlayer spacing did not change back. The simulated reaction rate constant was calculated from 0.002 min. -1 Increased to 0.0035min -1 The results are consistent with the experimental results (deviation ≤3%).

[0082] Example 6: Multi-system synergistic regulation verification Experimental design: Select CdS, TiO2, MoS2 three typical systems, set 3 parallel experiments for each group, respectively, adopt the method of the application (regulation group) and traditional experience regulation (control group). The regulation group strictly follows the S1-S5 steps, and the control group only adjusts a single parameter (such as light intensity) by artificial adjustment.

[0083] Regulation parameters: 1 General experimental platform Light source: 300 W xenon lamp, AM 1.5G filter, light intensity 100 mW cm -2 (point-by-point calibration ±3%).

[0084] Reaction cell: 100 mL high-purity quartz cell, constant temperature water bath ±0.1 ℃, back pressure valve to maintain total pressure 1.00 ±0.02 atm; O2≤ 5 ppm (online oxygen analyzer).

[0085] Raman: 532 nm, spot 100 μm, sampling frequency 2 Hz; automatic control through LabVIEW serial connection lamp power supply, mass flow controller and constant temperature bath.

[0086] Hydrogen production detection: GC-TCD, automatic sampling every 10 min, 6 times continuously (RSD ≤ 2%).

[0087] 2 CdS system (laser power regulation) Catalyst: 0.10 g CdS (30 nm, 0.5 wt% Pt photo-deposition promoter catalyst).

[0088] Solution: 100 mL 10 vol% methanol-water, Ar degassing for 30 min, flow rate 40 mL min -1 .

[0089] Temperature control: 25.0 ± 0.1 ℃.

[0090] Regulation logic: In-situ Raman 430 cm -1 peak shift amount ≥ 2.5 cm -1 → laser power 30 mW → 15mW (completed within 1 s); peak shift falls back <2 cm -1 Post-recovery 30 mW.

[0091] Results: Light corrosion rate 0.007 mg L -1 h -1(ICP-MS for Cd2+) and hydrogen production rate from 80 ± 2 to 129.6 ± 2.5 μιηοΙ g -1 h -1 (ICP-MS for Cd2+) and hydrogen production rate from 80 ± 2 to 129.6 ± 2.5 μιηοΙ g

[0092] 3 TiO2 system (hydrothermal temperature tuning) Catalyst: 0.15 g sol-gel anatase TiO2 (20 nm, 98% anatase).

[0093] Solution: 150 mL deionized water, N2 degassed for 30 min.

[0094] Pressure: total pressure 1 atm, N2 atmosphere.

[0095] Tuning logic: in-situ Raman I 144 / I 133 ≤ 1.2 (rutile phase ≥ 15%) → start constant bath temperature ramp: 5 °C min -1 → 180 °C, hold for 30 min; then 3 °C min -1 down to 25 °C, cycle 2 times.

[0096] Result: anatase retention ≥ 85%, hydrogen production rate from 90 ± 3 to 157.5 ± 4 μιηοΙ g - 1 h -1 (75% improvement).

[0097] 4 MoS2 system (interlayer spacing tuning) Catalyst: 0.50 g commercial MoS2 (original interlayer spacing 0.62 nm).

[0098] Intercalation: 50 mL 1.6 M n-butyllithium-hexanes, 25 °C stirring for 36 h under Ar protection.

[0099] Washing: dry hexanes wash 3 times after centrifugation.

[0100] Water exfoliation: wet cake was dropped into 100 mL deionized water, 200 W sonication for 20 min; XRD (002) peak was measured in situ, 2θ shifted to 11.8° (d = 0.75 nm) and stopped immediately; if over-exfoliated, 0.02-0.03 nm shrinkage could be recovered by 120 °C annealing for 1 h.

[0101] Reaction parameters: 0.08 g modified MoS2+ 80 mL 5 vol% lactic acid-water, N2 flow 30 mL min -1 , 25 °C.

[0102] Results: Interlayer spacing 0.75 nm, water molecule diffusion coefficient increased by 42%, hydrogen production rate from 75 ± 2 to 129.8 ± 3 μmol g -1 h -1 (73% increase).

[0103]

Control group example

[0104] Adjustment method: only manually rotate the xenon lamp power knob to change the light intensity, and the rest of the parameters are constant: initial light intensity 100 mW cm -2 ; manually increase by 20 mW cm -2 every 30 min, up to 200 mW cm -2 (total of 5 grades); no longer adjusted after reaching 200 mW cm -2 , run continuously to a total time of 6 h.

[0105] Record content: measure hydrogen production every 30 min with GC-TCD; Use the same Raman system and the same laser power (50 mW) to collect signals, only for post-comparison, not for feedback.

[0106] Termination condition; reaction 6 h, calculate the average hydrogen production rate; catalyst activity decreases by >50% to stop (Cd 2+ leaching 0.02 mg L -1 h -1 as the limit).

[0107] Number of repetitions: 3 times in parallel, take the average as the "single-parameter artificial adjustment" benchmark.

[0108] Result comparison: CdS hydrogen production rate of the regulation group increased from 80 to 129.6 μmol・g -1 ・h -1 (62% increase), and the control group only increased by 25%; TiO2 regulation group increased from 90 to 157.5 μmol・g -1 ・h -1(75%), the control group increased by 30%; the MoS2 regulation group increased from 75 to 129.8 muol.g -1 h -1 (73%), the control group increased by 28%.

[0109] In terms of reaction mechanism identification, the error of ·OH generation rate determined by coupling analysis of the regulation group is ≤4%, and the error of the control group is 15-20% due to the lack of simulation verification; in terms of catalyst stability, the activity retention rate of the regulation group CdS is 80% after 120h of continuous reaction, and that of the control group is reduced to 30% after 50h; in terms of TiO2 crystal phase transformation, the retention rate of anatase phase of the regulation group is stable at 85% within 100h, and that of the control group is reduced to 50% after 70h.

[0110] Different laboratory repeated experiments show that the relative standard deviation (RSD) of the regulation group is ≤5%, and the RSD of the control group is 12-18%, which verifies the reliability and universality of the method of the application. In summary, through the closed-loop system of "spectrum monitoring-simulation analysis-dynamic regulation", the average hydrogen production efficiency of the application is increased by 63% compared with the traditional method, and the regulation parameters can be quantified and the mechanism can be traced, which provides key support for the standardized application of photocatalytic hydrogen production technology.

[0111] The above shows and describes the basic principles, main features and advantages of the application. Those skilled in the art should understand that the application is not limited by the above examples, and the above examples and descriptions in the specification are only preferred examples of the application and are not intended to limit the application. Without departing from the spirit and scope of the application, various changes and improvements can be made to the application, and these changes and improvements all fall within the scope of the claimed application. The scope of protection of the application is defined by the appended claims and their equivalents.

Claims

1. A method for dynamically controlling the interfacial reaction in photocatalytic hydrogen production based on in-situ Raman spectroscopy-molecular dynamics coupled simulation, characterized in that, Includes the following steps: S1. Establishment of an in-situ Raman spectroscopy monitoring system: The system includes a transparent reaction cell equipped with a magnetically driven stirrer, a 532nm laser light source, a spectrometer, and a data acquisition device; the reaction cell has an inlet and an outlet, the inlet is connected to an inert gas pipeline and a hydrogen pipeline, and the outlet is connected to a gas chromatograph; the spectrometer has a resolution of not less than 1cm. -1 It is equipped with a cooled CCD detector with a cooling temperature of ≤-70℃; the data acquisition device is equipped with spectral analysis software to acquire Raman spectral data; S2. Constructing a molecular dynamics simulation model: Based on the Raman spectral data obtained in step S1, extract one or more of the catalyst crystal phase composition, lattice parameters, and bond length and bond angle information of adsorbed molecules using PeakFit software for peak fitting. Construct an initial molecular dynamics model containing the photocatalyst surface layer and the adsorbed water molecule layer using Materials Studio software. S3. Perform molecular dynamics simulations: The ab initio molecular dynamics AIMD simulation method was adopted, using the VASP 6.3.0 software package. The exchange correlation functional was PBE+D3, the cutoff energy was set to 400-500 eV, and the K-point sampling used a Γ-point grid. The interatomic interactions were described by the projected plane wave method. The simulation time step was 1-2 fs, and the total simulation time was not less than 100 ps. S4. Coupling Analysis Data: Using a data fusion algorithm combining principal component analysis for dimensionality reduction and partial least squares regression, the Raman spectroscopy data from step S1 and the molecular dynamics simulation results from step S3 are correlated, and a three-dimensional response surface model is constructed using Origin software; S5. Dynamic control of the reaction: Based on the results of the coupling analysis, the photocatalytic hydrogen production is dynamically controlled.

2. The method according to claim 1, characterized in that, In step S1, both the inlet and outlet of the reaction tank are connected in series with glass rotor flow meters and precision needle valves. The catalyst and reaction solution are placed in the reaction tank. Before the reaction, inert gas is introduced through the inlet to remove air. Then, a mixture of inert gas and hydrogen is introduced to simulate the reaction environment under hydrogen partial pressure conditions. The outlet is connected to a gas chromatograph to monitor the hydrogen production in real time. The sampling frequency of the data acquisition device is controlled at 0.1-10Hz to monitor the photocatalytic hydrogen production process in real time and obtain Raman spectral data.

3. The method for dynamic control of the interfacial reaction in photocatalytic hydrogen production based on in-situ Raman spectroscopy-molecular dynamics coupling simulation according to claim 1, characterized in that, In step S1, the light-transmitting reaction tank is made of quartz glass with a purity of ≥99.9%, and the tank body is provided with an air inlet and an air outlet with an inner diameter of 3-5mm. And / or, the impeller blades are made of polytetrafluoroethylene, and the stirring speed can be steplessly adjusted within the range of 100-500 r / min; And / or, the laser source is an Nd:YAG solid-state laser, the power is continuously adjustable in the range of 10-500mW by the electronic control module, and the laser spot diameter is controlled in the range of 50-200μm by the focusing lens group; And / or, the spectrometer employs a backscattering optical path design, with a detection range covering 200-4000 cm⁻¹. -1 ; And / or, the acquired Raman spectral data include the 400-500 cm⁻¹ region corresponding to the lattice stress changes induced by CdS photocorrosion. -1 Characteristic peak shift, TiO2 phase transformation related to 144 cm⁻¹ -1 Anatase and 133cm -1 The intensity variation of rutile characteristic peaks, and the ·OH intermediate at 3000-3600 cm⁻¹ -1 The characteristic peaks and OH bond vibration frequency shifts are any one or more of the following:

4. The method for dynamically controlling the interfacial reaction in photocatalytic hydrogen production based on in-situ Raman spectroscopy-molecular dynamics coupling simulation according to claim 1, characterized in that, The thickness of the photocatalyst surface layer in S2 is 2-5 nm, containing 3-5 atomic layers; the number of adsorbed water molecules in the layer is 50-200, calculated based on the actual water coverage. And / or, in step S2, the model's x, y, z dimension ratio is 1:1:(2-3), the boundary adopts three-dimensional periodic boundary conditions, the simulation system temperature is controlled by Nose-Hoover, and the pressure is controlled to 0.95~1.05 atm by the Parrinello-Rahman method.

5. The method for dynamic control of the interfacial reaction in photocatalytic hydrogen production based on in-situ Raman spectroscopy-molecular dynamics coupling simulation according to claim 1, characterized in that, The AIMD simulation in step S3 also includes the calculation of the separation efficiency of photogenerated electron-hole pairs. The Bader charge analysis method combined with electronic locality function analysis is used to determine the charge transfer amount and carrier lifetime. During the simulation, time-resolved fluorescence spectroscopy data is used to calibrate the carrier dynamics parameters.

6. The method for dynamically controlling the interfacial reaction in photocatalytic hydrogen production based on in-situ Raman spectroscopy-molecular dynamics coupling simulation according to claim 1, characterized in that, Step S4 also includes using Origin software to construct a three-dimensional response surface model to determine one or more of the following: the generation rate of reaction intermediates ·OH and / or H2O2, the time points of chemical bond breaking / formation, and the activation energy of the rate-determining step of the reaction.

7. The method for dynamically controlling the interfacial reaction in photocatalytic hydrogen production based on in-situ Raman spectroscopy-molecular dynamics coupling simulation according to claim 1, characterized in that, The parameters dynamically controlled in step S5 include any one or more of the following: photocatalyst composition, interlayer spacing of layered photocatalyst, reaction atmosphere, and reaction temperature (25-80℃).

8. The method for dynamic control of the interfacial reaction in photocatalytic hydrogen production based on in-situ Raman spectroscopy-molecular dynamics coupling simulation according to claim 1, characterized in that, In step S1, the dedicated software on the data acquisition device has a real-time plotting function, used to dynamically display the curve of Raman characteristic peak intensity changing with reaction time, and can automatically identify the characteristic peak drift trend. When CdS is detected at 430 cm⁻¹, -1 Characteristic peak shift exceeds 5cm -1 It triggers an early warning of photocorrosion and automatically adjusts the power of the light source through a software interface.

Images