Impurity control methods in the preparation of sodium methoxide

Abstract

This invention discloses an impurity control method in the preparation of sodium methoxide, relating to the field of organometallic compound preparation and purification technology. The method includes the following steps: S1, collecting temperature data from multiple points on the surface of a membrane module, constructing a temperature distribution matrix, and identifying high-risk areas with a risk of crystallization; S2, for the identified high-risk areas, adjusting the local frequency output of a vibration device to generate directional spatial focusing perturbations in the target area, thereby achieving local temperature perturbation intervention. This invention achieves high-resolution visualization of the membrane surface state by constructing and updating temperature and concentration matrices. Combined with spatial focusing perturbations and reverse pulse intervention, it actively interrupts crystal nucleus formation, reducing the risk of membrane blockage. Furthermore, based on feedback dynamic parameter correction, it constructs a closed-loop control system, improving the membrane module's anti-crystallization capability and operational stability, demonstrating significant industrial application value.

Description

Technical Field

[0001] This invention relates to the field of organometallic compound preparation and purification technology, specifically to a method for controlling impurities in the preparation process of sodium methoxide. Background Technology

[0002] Impurity control in the preparation of sodium methoxide refers to the technical means of suppressing or removing non-target substances during the sodium methoxide formation reaction and its post-processing by optimizing reaction conditions, adjusting process parameters, or introducing specific impurity removal measures. Sodium methoxide is usually produced by reacting metallic sodium with anhydrous methanol. However, in actual production, impurities such as moisture, air, and carbon dioxide are often introduced, leading to the formation of byproducts such as sodium hydroxide, sodium carbonate, hydrates, or unreacted residual methanol. These impurities significantly affect the purity, storage stability, and catalytic or synthetic efficacy of sodium methoxide in subsequent reactions. Therefore, impurity control not only includes preventing the formation of impurities but also encompasses the removal of generated impurities through separation, washing, drying, vacuum treatment, and other methods, ultimately achieving the stable preparation of high-purity sodium methoxide that meets the activity and safety requirements for industrial applications.

[0003] In the sodium methoxide preparation process, dynamic vibration membrane separation technology is introduced as one of the impurity control methods and applied in the post-processing stage to achieve fine separation and purity improvement of the product. This method uses a high-frequency vibration-driven nanoscale ceramic membrane, which can efficiently retain suspended particles, unreacted byproducts, crystallized salts, and trace colloidal impurities in the liquid phase system, and has excellent selective filtration capabilities.

[0004] Compared to traditional static membrane filtration, dynamic vibration can significantly reduce membrane surface fouling and flux attenuation, improve the operational stability and efficiency of the filtration system, and avoid secondary pollution or product loss caused by membrane clogging. This method is particularly suitable for high-purity applications, such as catalyst preparation and pharmaceutical intermediate synthesis. It can achieve continuous, physical, and high-precision removal of impurities while ensuring that the activity of the sodium methoxide main product is not affected, significantly improving process consistency and product quality stability.

[0005] The existing technology has the following shortcomings:

[0006] In the process of controlling impurities in sodium methoxide solution using dynamic vibration membrane separation, the membrane module is under continuous high-frequency vibration, which easily creates microscale disturbance flow fields in local areas. This, coupled with the instantaneous temperature rise caused by vibration friction, leads to fluctuations in the thermodynamic stability of the solution near the membrane surface. When the local solution concentration exceeds the solubility limit of sodium methoxide within a short period, resulting in instantaneous supersaturation, sodium methoxide crystals are likely to rapidly precipitate and accumulate on the membrane surface, forming a "crystallization burst zone." The large number of rapidly generated crystals in this zone easily connect laterally between membrane pores, forming crystal bridging structures. This causes the filtration pores to be completely blocked, making it impossible to maintain normal permeate flux. This type of "bridging" blockage is usually irreversible, making it difficult to restore membrane function through conventional methods such as backflushing and cleaning. This can lead to sudden interruption of the filtration process, system failure, and even contamination and scrapping of the entire batch of products. Summary of the Invention

[0007] The purpose of this invention is to provide an impurity control method in the preparation process of sodium methoxide. By constructing and updating the temperature and concentration matrix, high-resolution visualization of the membrane surface state is achieved. Combined with spatial focusing perturbation and reverse pulse intervention, crystal nucleus generation is actively interrupted, reducing the risk of membrane blockage. Based on feedback dynamic correction parameters, a closed-loop control system is constructed to improve the anti-crystallization ability and operational stability of the membrane module, which has significant industrial application value.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a method for controlling impurities in the preparation process of sodium methoxide, comprising the following steps:

[0009] S1. Collect temperature data at multiple points on the surface of the membrane module, construct a temperature distribution matrix, and identify high-risk areas with the risk of crystallization precipitation;

[0010] High-sensitivity micro thermocouple temperature sensors are deployed on the surface of the membrane module to collect multi-point temperature data in real time. A two-dimensional interpolation algorithm is used to generate and visualize the temperature distribution matrix. Based on the conditions that the local temperature is 0.8 degrees Celsius higher than the overall average, or the temperature difference between adjacent areas is greater than 5 degrees Celsius per centimeter, or the temperature rises continuously for more than 2 minutes without a downward inflection point, high-risk areas for crystallization are identified and their coordinates and temperature change trends are recorded.

[0011] S2, for the identified high-risk areas, adjust the local frequency output of the vibration device to generate directional spatial focusing disturbance in the target area, and realize local temperature disturbance intervention;

[0012] S3 induces microscale turbulence in the liquid phase under the effect of spatial focusing disturbance, dynamically homogenizes the solute concentration in the target area, and simultaneously collects concentration data to construct an initial concentration distribution matrix;

[0013] S4. Based on the initial concentration distribution matrix, identify regions where abrupt concentration gradients still exist, and dynamically update the concentration distribution matrix during the turbulence averaging process to extract abnormal concentration regions and determine the location of reverse intervention.

[0014] S5 applies short-cycle, high-energy reverse pulse vibrations to the abnormal concentration region of the extract, thereby disrupting the initial crystal nucleus structure and inhibiting crystal bridging and membrane pore blockage.

[0015] S6, based on the trend of temperature and concentration matrix changes after reverse intervention, jointly analyzes the response of each high-risk area, dynamically corrects the vibration frequency and temperature control strategy, and achieves closed-loop precise control of the crystal precipitation process.

[0016] Preferably, the specific steps for using a two-dimensional interpolation algorithm to compensate for the temperature between sampling points and generate a complete temperature distribution matrix are as follows:

[0017] The surface of the membrane module is established as a two-dimensional coordinate plane, and each temperature acquisition point is mapped to a two-dimensional coordinate point with a fixed position and assigned a corresponding temperature value.

[0018] Choose one of the interpolation methods, such as bilinear interpolation, bicubic interpolation, or Gaussian kernel regression interpolation, and calculate the temperature value at the unknown location based on the spatial relationship between known points to achieve a smooth transition and continuous distribution of the temperature field.

[0019] Based on the actual size of the membrane module, the membrane surface is divided into a fine grid, and the temperature value of each grid point is calculated using an interpolation function to construct a temperature distribution matrix containing temperature data from all locations.

[0020] Preferably, step S2 includes:

[0021] Two-dimensional coordinates and temperature change trends are extracted from high-risk areas identified in the temperature distribution matrix.

[0022] Input the coordinates of the high-risk area into the vibration control unit, call the pre-established vibration frequency response model, and calculate the optimal frequency output value and vibration energy parameters;

[0023] The driving vibration device generates spatial focusing disturbance above the target area, which acts on the liquid boundary layer to disrupt local thermal accumulation and concentration retention.

[0024] Monitor the temperature change trend of the target area. If the temperature tends to stabilize, maintain the current frequency output; otherwise, recalculate and adjust the frequency parameters.

[0025] Preferably, step S3 includes:

[0026] Based on the location of high-risk areas, the vibration device outputs spatial focusing disturbances with wavelengths of 1 to 10 millimeters and amplitudes of 10 to 500 micrometers to form a microscale turbulent flow field in the liquid phase boundary layer of the membrane surface;

[0027] During the turbulence process, high-precision concentration sensors are used to simultaneously collect solute concentration data at multiple locations in the target area and perform normalization processing.

[0028] The collected concentration data is mapped to the two-dimensional coordinate plane of the membrane module, and a two-dimensional interpolation algorithm is used to estimate the concentration values ​​between sampling points;

[0029] An initial concentration distribution matrix is ​​generated according to the membrane surface grid division rules to reflect the solute distribution state of the liquid phase on the membrane module surface.

[0030] Preferably, step S4 includes:

[0031] The initial concentration distribution matrix is ​​divided into a two-dimensional grid, and the concentration gradient between each grid point and its adjacent points is calculated.

[0032] Based on the set concentration gradient threshold, identify abrupt change regions and determine whether there is a concentration change trend that concentrates inward.

[0033] During the turbulence homogenization process, concentration data is collected at fixed intervals and the concentration distribution matrix is ​​dynamically updated.

[0034] Regions that maintain a high concentration gradient over multiple consecutive cycles are selected as target locations for reverse vibration intervention.

[0035] Preferably, step S5 includes:

[0036] Abnormal concentration gradient regions are located and intervention priorities are set based on dynamically updated concentration distribution matrices;

[0037] Short-period, high-energy reverse pulse vibrations are applied to the target region to drive the liquid phase to generate asymmetric perturbation eddies.

[0038] By using pulsed kinetic energy to disrupt the aggregated structure of solute particles, the crystal nucleation path is blocked and crystal bridging is prevented.

[0039] After intervention, continue to collect concentration data and determine the trend of concentration gradient changes to decide whether to carry out subsequent intervention operations.

[0040] Preferably, step S6 includes:

[0041] Collect temperature and concentration data after reverse pulse vibration intervention and update the temperature distribution matrix and concentration distribution matrix;

[0042] Based on the trend of matrix changes, a response characteristic judgment model is established and the intervention effect levels are classified;

[0043] The frequency output parameters and temperature control strategy of the vibration device are dynamically adjusted according to the response level.

[0044] The corrected parameters will be input into the control process, and a new round of monitoring, judgment, and control operations will be executed.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] This invention overcomes the limitations of traditional membrane separation processes, which rely solely on post-processing intervention and lack dynamic prediction capabilities, by real-time construction and updating of temperature and concentration distribution matrices. It achieves, for the first time, high-resolution visualization of membrane surface thermodynamics and substance concentration states, providing a scientific basis for risk identification. Furthermore, based on the spatial positioning of high-risk areas, it employs a localized, precise control method combining spatial focusing perturbation and reverse pulse intervention. This proactively intervenes before crystals stabilize, breaking the initial crystal nucleus structure and significantly reducing the risks of crystal bridging and membrane blockage. Simultaneously, through feedback analysis of the intervention results, it dynamically corrects the vibration frequency and temperature control strategy, constructing a closed-loop control system of "data-driven—parameter adaptive—process steady-state maintenance." Overall, this invention not only improves the anti-crystallization capability and service life of membrane modules but also effectively reduces cleaning frequency and system downtime risks while ensuring separation purity and continuous operation, demonstrating outstanding industrial adaptability and application value. Attached Figure Description

[0047] Figure 1 This is a flowchart of the impurity control method in the preparation process of sodium methoxide according to the present invention. Detailed Implementation

[0048] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0049] This invention provides a method for controlling impurities during the preparation of sodium methoxide, comprising the following steps:

[0050] S1. Collect temperature information from multiple points on the surface of the membrane module, construct a complete temperature distribution matrix on the surface of the membrane module, and identify high-risk areas on the surface of the membrane module with the risk of crystallization based on the temperature distribution matrix, so as to provide a data basis for vibration intervention and control of target areas;

[0051] To address the potential risk of localized crystallization of sodium methoxide solution during dynamic vibration membrane filtration, a temperature distribution matrix is ​​constructed to identify high-risk areas on the membrane module surface. The specific steps include:

[0052] During the operation of the filtration equipment, multiple temperature sensors are evenly distributed on the surface of the membrane module to collect surface temperature information at various points in real time. The temperature sensors used are high-sensitivity miniature thermocouples with a detection accuracy better than ±0.05℃. The spacing between them is controlled within the range of 1–3 cm to ensure coverage of the temperature field changes across the entire membrane module surface. To avoid interference from the sensors on the fluid flow, each sensor is tightly attached to the membrane module surface via a flexible sensing patch with high thermal conductivity but low interference, and is connected to the data acquisition device via a shielded cable to achieve high-frequency (not less than 10 Hz) temperature data acquisition.

[0053] The collected multi-point temperature data is transmitted to the data processing module in real time, and a two-dimensional interpolation algorithm is used to smooth the transition between temperature values ​​of the sampling points to generate a complete and continuous temperature distribution matrix. This temperature distribution matrix dynamically displays the real-time temperature changes of various regions on the surface of the membrane module in the form of a digital heatmap, thus providing a visual representation of the thermodynamic state of the membrane surface at the microscale. During the temperature data processing, a gradient change threshold recognition mechanism is specifically set up to capture local points or small-scale clusters of temperature abrupt changes in the membrane surface region.

[0054] "Two-dimensional interpolation algorithm" refers to using mathematical methods to calculate the values ​​of unknown positions between certain discrete points (such as temperature) on a two-dimensional plane (e.g., the surface of a membrane module), thereby achieving continuous distribution and smooth transition of data throughout the entire area. In the sodium methoxide preparation impurity control method of this invention, the function of this algorithm is to expand and compensate the temperature data of multiple fixed measuring points on the surface of the membrane module into a continuous temperature distribution map, i.e., a "temperature distribution matrix," on the entire membrane surface.

[0055] Specifically, the role of two-dimensional interpolation is to eliminate the spatial measurement gap caused by the limited number of sensors, so that the temperature field presents a more realistic, continuous and analyzable state, thereby identifying high-risk areas for crystal precipitation such as microscale temperature rise regions and temperature gradient abrupt change regions.

[0056] The following are the specific steps for generating a temperature distribution matrix using two-dimensional interpolation:

[0057] The surface of the membrane module is defined as a two-dimensional coordinate plane, and each temperature acquisition point is mapped to a two-dimensional coordinate point with a fixed position. Each sampling point corresponds to a temperature value, forming a discrete set of point data.

[0058] Choose an interpolation method suitable for strong physical continuity, such as bilinear interpolation, bicubic interpolation, or Gaussian kernel regression interpolation. The interpolation function uses the known temperature at the measurement point as a basis and calculates the temperature value at any position between two known points according to interpolation rules, ensuring the continuity and smoothness of the transition process. For example, bilinear interpolation derives the temperature of intermediate unknown points by performing linear estimations sequentially along the x-axis and y-axis.

[0059] Based on the actual dimensions of the membrane module, the entire membrane surface is divided into a fine grid. The temperature value of each grid point is estimated using an interpolation function, ultimately constructing a complete two-dimensional temperature distribution matrix. The value at each position in this matrix is ​​an estimated temperature, reflecting the local thermal distribution state of the membrane surface.

[0060] The temperature matrix is ​​visualized as a pseudo-color thermal map or contour map, allowing operators or algorithm control devices to intuitively identify areas of sudden temperature changes or local overheating, thus serving as important reference data for subsequent identification of high-risk areas for crystallization and adjustment of vibration parameters.

[0061] In summary, the core value of two-dimensional interpolation algorithms lies in generating high-density, high-continuity, and high-reliability compensated temperature field images in space using limited actual measurement point data, thereby supporting accurate identification and control judgment of the membrane surface state.

[0062] Based on the constructed temperature distribution matrix, real-time analysis of the membrane module surface is performed to identify high-risk areas with potential crystallization precipitation. The criteria for identifying high-risk areas are defined as three independent physical conditions; if any one of these conditions is met, the area is considered to have the potential for initial crystal precipitation. First, when the temperature at a local location on the membrane module surface is more than 0.8 degrees Celsius higher than the overall average temperature of the membrane surface under current operating conditions, heat accumulation is identified at that location. This heat accumulation may reduce the solubility of the solute in the local liquid phase, thereby increasing the risk of initial crystal precipitation. Second, when the temperature change between two adjacent locations on the membrane module surface exceeds 5 degrees Celsius per centimeter, a significant temperature gradient abrupt change is identified at that location. This abrupt change may trigger changes in local thermal convection and solute migration direction, thereby promoting rapid precipitation of the solute in the lower-temperature region. Third, if the temperature at a certain location on the surface of the membrane module maintains an upward trend for more than 2 minutes during continuous monitoring, and the temperature increase does not show a decrease or a stable inflection point during this period, it is determined that there is continuous heat input or insufficient heat dissipation at that location. This thermal condition will cause the local liquid phase to reach or exceed the supersaturation critical point of the solute, inducing crystal nucleation. Once any one of the above three conditions is met, the location can be designated as a high-risk area for crystallization, and subsequent vibration intervention steps can be initiated to prevent rapid crystal growth and structural bridging. Furthermore, by retrospectively comparing historical temperature evolution data, the fluctuation frequency and duration of the high-risk area can be determined, providing a dynamic risk level assessment for subsequent intervention strategies.

[0063] The identified high-risk areas are used as target areas for temperature control and disturbance intervention. Parameters such as the coordinates of the identified areas, temperature change trends, and gradient intensity are recorded and combined with subsequent vibration frequency adjustment and direction focusing logic to construct a temperature-driven real-time dynamic control path. This path not only supports the prediction and calibration of initial crystallization trends but also serves as input for crystal precipitation suppression measures.

[0064] This step serves to provide precise and quantifiable criteria for subsequent crystallization risk intervention, and is a crucial preliminary step for achieving efficient impurity control in the dynamic vibration membrane separation process. In the post-processing of sodium methoxide preparation, the membrane module is under high-frequency vibration for extended periods, leading to transient temperature rises in localized areas. This fluctuation in solution thermodynamic stability can easily induce rapid local crystal precipitation and the formation of bridging and blocking structures between membrane pores. By deploying multiple temperature acquisition points on the membrane module surface and acquiring real-time temperature data, the thermal change trends in various regions of the membrane surface can be effectively perceived. Furthermore, using temperature interpolation and matrix calculation methods, the limited point data can be expanded into a continuous two-dimensional temperature distribution map, achieving a comprehensive representation of the membrane surface thermal state. Based on this temperature distribution matrix, potential risk points such as areas of abnormal temperature rise and abrupt temperature gradient changes can be identified, thereby determining high-risk areas that may induce crystallization. These areas, as target locations for vibration control, can guide the precise adjustment of subsequent vibration frequency, disturbance mode, and intervention energy, ensuring highly targeted and timely intervention measures. Therefore, this step not only realizes the transformation from "passive monitoring" to "active prediction", but also provides a data foundation and spatial positioning support for establishing a temperature-driven closed-loop vibration control mechanism, which is a key technical prerequisite for improving the stability of membrane separation process and product purity.

[0065] S2, based on the high-risk areas identified in the temperature distribution matrix, adjust the local frequency output parameters of the vibration device in the corresponding areas to make the vibration device generate directional spatial focusing disturbances in the high-risk areas, so as to implement local disturbance intervention in the temperature fluctuation areas.

[0066] For high-risk areas of membrane module surface crystallization identified based on the temperature distribution matrix, in order to effectively avoid the formation of crystal bridging structures caused by local oversaturation, directional spatial focusing disturbance intervention is achieved in the target area by adjusting the local frequency output of the vibration device. The specific steps include:

[0067] Based on the constructed temperature distribution matrix, and combined with the set temperature rise threshold, temperature gradient change threshold, and thermal stability deviation index, point-by-point analysis is performed on the surface of the membrane module to screen out local areas with a tendency to crystallize, and extract their two-dimensional coordinate information, temperature change trend, and risk level. Through this process, multiple potential crystallization risk points can be accurately marked on the membrane surface, laying the foundation for subsequent spatial positioning and parameter setting for intervention.

[0068] Based on the identified high-risk area locations, the coordinate data of these areas is input into the vibration control unit, triggering the calculation of local frequency parameters for the corresponding areas. During this process, based on the thermal disturbance characteristics of the high-risk areas and a pre-established vibration frequency response model, the optimal frequency output value and vibration energy parameters applicable to the area are calculated. This model comprehensively considers local liquid phase concentration gradients, fluid viscosity changes, heat conduction rates, and vibration waveform propagation characteristics, ensuring that the output frequency has a high degree of adaptability to the target risk area.

[0069] A "pre-established vibration frequency response model" refers to a mathematical or engineering model constructed before implementation through theoretical analysis, experimental data, or simulation calculations. This model describes the response behavior of the membrane module surface under different vibration frequencies. Specifically, by pre-testing the disturbance effects of different frequencies, amplitudes, and waveform parameters on the liquid phase membrane surface region, the model establishes a functional relationship between frequency and multiple response variables, such as local liquid disturbance intensity, heat transfer efficiency, and temperature change rate. When targeted intervention is needed in a high-risk area of ​​the membrane surface, this model can be used to quickly calculate the most suitable frequency output value for that area. This ensures that vibration acts on the liquid phase boundary layer at the most effective location and intensity, breaking up local supersaturation or stagnant zones and inhibiting crystal nucleation and accumulation. This model transforms the vibration parameter adjustment process from experience-dependent to data-driven, improving the accuracy, stability, and response speed of intervention. It is a fundamental support for achieving precise vibration control.

[0070] Based on the calculated frequency parameters, an actuator at a predetermined position is driven in the vibration device to generate spatially focused disturbances above or near the corresponding high-risk areas on the membrane module surface. These spatially focused disturbances, through mechanical vibrations of specific wavelengths and amplitudes, act on the boundary layer between the liquid phase and the membrane surface, causing a stable and uniform turbulent flow state in that region. This breaks the supersaturated boundary conditions caused by localized heat accumulation or fluid stagnation, thereby reducing the probability of initial crystal nucleation.

[0071] Spatial focusing perturbation refers to the use of mechanical vibrations with specific wavelengths and amplitudes to concentrate energy on a localized area of ​​the membrane module surface, creating a regular, turbulent flow within the boundary layer between the liquid phase and the membrane surface in that area. This perturbation disrupts the stable boundary layer structure that might otherwise be formed due to reduced liquid flow rate, heat accumulation, or restricted mass migration. Once the boundary layer stabilizes, the local solution temperature rises, and the solute concentration increases, easily leading to a "supersaturated" state and inducing initial crystal nucleation. Spatial focusing perturbation, by enhancing local mixing and promoting the diffusion of heat and concentration, allows heat to be carried away promptly and the concentration to be dynamically homogenized, effectively disrupting the preconditions for supersaturation and reducing the probability of crystal formation in that region from the outset. This method is more targeted and efficient than traditional overall stirring, and is a key means of controlling the risk of precise local crystallization while maintaining overall stability.

[0072] During the implementation of space-focused perturbations, the effect of vibration intervention on the thermal state of the membrane surface was verified by continuously monitoring the temperature change trend of the target area. Once the temperature gradient was significantly alleviated and the regional temperature tended to reach a stable equilibrium, the current frequency output was deemed effective. If the temperature anomaly persisted, the frequency parameters and vibration duration were automatically reassessed and dynamically adjusted until the thermodynamic state of the target area returned to a safe range. This entire process, through a combination of continuous feedback and targeted intervention, constructed a predictive and adaptive space perturbation intervention path, significantly improving the response capability and suppression efficiency of vibration control to the risk of localized crystallization on the membrane surface.

[0073] This step aims to achieve precise intervention and dynamic control over high-risk areas of localized crystallization on the membrane module surface, thereby effectively suppressing the unwanted crystallization of sodium methoxide on the membrane surface and preventing the formation of crystal bridging structures and clogging of the filtration channels. During the dynamic membrane separation process of sodium methoxide solution, certain local areas are prone to forming micro-regions with abnormally high temperatures due to vibration friction, heat accumulation, or fluid stagnation. This leads to a decrease in the thermodynamic stability of the solution, inducing short-term supersaturation and becoming potential trigger points for rapid crystal precipitation. To address this issue, this step identifies these high-risk areas through real-time analysis of the temperature distribution matrix. Based on this, the frequency output parameters of the vibration device are adjusted to generate directional, concentrated, and energy-controllable "spatial focusing disturbances" within the identified high-risk areas. These disturbances break the flow stagnation caused by thermal gradients within the membrane boundary layer, promoting the formation of microscale turbulent flow fields in the local liquid phase, thereby improving local heat and mass transfer efficiency, homogenizing temperature and concentration distribution, and suppressing the formation of crystallization conditions. Unlike traditional overall vibration or static operation, this step realizes the control strategy of "intervention on demand, fixed-point disturbance, and real-time response", which significantly improves the control accuracy of the microscopic physical state of the membrane surface and is one of the core technical links to ensure continuous operation stability and separation effect.

[0074] S3 utilizes spatial focusing perturbation to induce a microscale turbulent field in the liquid phase on the surface of the membrane module. The turbulence effect dynamically homogenizes the liquid phase concentration in high-risk areas. Simultaneously, during the perturbation process, solute concentration information of the liquid phase in the target area is collected to construct the initial concentration distribution matrix on the surface of the membrane module.

[0075] To achieve precise intervention and status monitoring of high-risk areas for crystallization of sodium methoxide solution on the membrane module surface, microscale turbulence is induced by spatial focusing perturbation, and a membrane surface concentration distribution matrix is ​​constructed by combining real-time concentration acquisition methods. The specific steps include:

[0076] In high-risk areas of crystallization on the identified membrane module surface, the frequency output of the vibration device is adjusted to generate directional and localized spatially focused disturbances. These disturbances propagate vertically or tangentially to the liquid boundary layer along the target area with specific wavelengths and amplitudes. This focused disturbance creates a highly confined microscale disturbance flow field near the membrane surface, with its disturbance scale controlled within the sub-millimeter range. This effectively disrupts the static boundary layer structure in the liquid phase near the membrane surface, enhances the local fluidity of the liquid in this region, and strengthens the convection and diffusion of the solute. Unlike traditional large-scale stirring or overall vibration methods, spatially focused disturbances, through concentrated energy delivery, avoid overall flow field turbulence while efficiently altering the local solution state.

[0077] "Specific wavelength and amplitude" refers to two key parameters of the mechanical vibration propagating in the liquid phase when spatial focusing perturbation is performed: wavelength represents the periodic distance of the vibration propagation in space, and amplitude represents the maximum displacement or energy intensity of the vibration. Together, they determine the range and intervention capability of the perturbation in the liquid phase boundary layer of the membrane.

[0078] The specific wavelength is predetermined based on the physical properties of the liquid phase (such as density, viscosity, and temperature) and the energy transfer efficiency between the film surface and the liquid phase. The aim is to create a stable, periodic disturbance structure in a localized region, concentrating the turbulence without causing widespread chaos. In this technology, the wavelength is typically set between 1 mm and 10 mm. Shorter wavelengths help create high-frequency turbulence within small areas, thereby achieving more precise local concentration homogenization and thermal diffusion regulation.

[0079] The specific amplitude represents the displacement or energy intensity generated by the vibration, directly affecting the penetration depth and turbulence intensity of the disturbance. Too small an amplitude may not effectively affect the liquid boundary layer, while too large an amplitude may cause disturbance or even damage to the film surface. Therefore, the amplitude is generally controlled within the range of 10 micrometers to 500 micrometers, adjusted through experimental measurements and feedback from operating parameters to ensure that the disturbance sufficiently disrupts the boundary layer structure without causing liquid splashing or film surface resonance damage.

[0080] By precisely setting these two parameters, spatial focusing perturbations can act vertically or tangentially on the liquid boundary layer, forming stable, continuous, and directionally controlled micro-perturbations within the target region, thereby achieving physical suppression of local crystallization trends and concentration field homogenization.

[0081] Under the influence of spatial focusing perturbation, periodic disturbances and micro-swirling structures appear within the liquid phase of high-risk regions on the membrane surface. This breaks down the concentration gradient caused by local static accumulation, promoting dynamic homogenization of the solute within a microscopic range. This process significantly weakens the supersaturation tendency in local areas and reduces the probability of initial crystal nucleation. Especially in the early stages when temperature changes are not yet significant, concentration homogenization can serve as a preventative measure, inhibiting the formation of a suitable crystal growth environment in advance. The intensity and frequency of the perturbation generated by this micro-perturbation can be precisely controlled by adjusting the vibration waveform and driving cycle, making the concentration homogenization process stable, controllable, and sustainable.

[0082] While the micro-perturbation process continues, multiple high-precision liquid phase concentration sensors are installed on the surface of the membrane module to collect real-time solute concentration information within the target area. The concentration sensors employ miniature optical refractive index detection units or micro-conductivity probes, with a response time of less than 1 second and a measurement error of less than ±2%. Each sensor is precisely positioned to correspond to the perturbation area and records concentration data at a high frequency, which is then digitally stored via a data acquisition device. To ensure data accuracy, the acquisition period is synchronized with the perturbation cycle, and a filtering algorithm is used to remove high-frequency noise and background fluctuations unrelated to the perturbation.

[0083] The collected concentration data from each measuring point were normalized according to their corresponding spatial location, and a complete initial concentration distribution matrix was generated using a two-dimensional interpolation algorithm. This matrix uses the membrane module surface as a two-dimensional coordinate plane, with each cell corresponding to a solute concentration value at a given time, reflecting the concentration changes of the solution at the membrane surface at a microscale. This initial concentration distribution matrix can not only be used to determine whether the current concentration is in a homogenized state, but also provides fundamental data support for subsequent identification of local concentration abrupt changes, analysis of concentration perturbation effects, and implementation of reverse intervention strategies.

[0084] A complete initial concentration distribution matrix is ​​generated using a two-dimensional interpolation algorithm. This primarily utilizes solute concentration data obtained from multiple fixed sampling points on the membrane module surface to calculate the concentration values ​​of all unmeasured areas between these sampling points, thus achieving a continuous expression of the membrane surface solution concentration across the entire two-dimensional space. Specifically, firstly, a two-dimensional coordinate plane is established on the membrane module surface (e.g., the x-axis represents the length of the membrane surface, and the y-axis represents the width), and the positions of all known sampling points and their corresponding concentration values ​​are input into this coordinate system. Next, an appropriate interpolation algorithm (such as bilinear interpolation, bicubic interpolation, radial basis function interpolation, etc.) is selected, and based on the spatial distance between each unknown point and surrounding known points and the concentration gradient trend, the most probable concentration value at that location is calculated, thereby filling in the blank areas of the concentration data. The entire membrane surface is divided into a fine grid, and each grid point is assigned a concentration value, ultimately forming a two-dimensional numerical matrix composed of horizontal and vertical coordinates and corresponding concentrations, which is the initial concentration distribution matrix. This matrix fully expresses the liquid phase solute distribution state at each location on the membrane module surface, providing fundamental data support for identifying local concentration anomalies, analyzing concentration homogenization effects, and implementing subsequent precise interventions.

[0085] This step aims to actively induce microscale turbulence in high-risk areas on the membrane module surface by introducing controllable local perturbations. This breaks the static boundary layer structure that may form in the liquid phase, achieving dynamic homogenization of local solute concentration. Concentration data is simultaneously collected during this process to construct an initial concentration distribution matrix reflecting the solute distribution on the membrane surface. In the dynamic separation of sodium methoxide, high-concentration solutions are prone to localized enrichment on the membrane surface due to temperature rise, stagnation, or insufficient perturbation, leading to short-term supersaturation and inducing initial crystal nucleation. Traditional separation methods struggle to identify and intervene in microscale concentration changes on the membrane surface in real time, easily resulting in delayed response or complete omission of local crystallization risks. This step adjusts the frequency output of the vibration device to generate spatially focused perturbations in the identified high-risk areas. The focused perturbation energy acts on the boundary layer between the liquid phase and the membrane surface, forming a stable microscale turbulence field. This improves the convective mass transfer efficiency in this area, allowing the local solute to redistribute and reach equilibrium. Under disturbance, real-time concentration data from multiple points within the area are simultaneously collected using high-precision concentration sensors and integrated into a complete initial concentration distribution matrix through spatial interpolation algorithms. This matrix not only serves as the basis for judging the concentration homogenization effect and assessing the crystallization risk, but also provides spatial reference for subsequent abnormal area identification and vibration intervention strategy optimization. In summary, this step achieves the synergistic goal of integrating information acquisition and risk control while ensuring stable solute distribution, and is a crucial component of this invention that deeply integrates physical disturbance with data-driven analysis.

[0086] S4. Based on the initial concentration distribution matrix, identify local areas with abrupt concentration gradients. While the turbulence homogenization process continues, dynamically update the concentration distribution matrix to extract areas with abnormal concentration gradients after homogenization, which are used to determine the target location for reverse vibration intervention.

[0087] Based on the initial concentration homogenization achieved through spatial focusing perturbation, the initial concentration distribution matrix is ​​continuously analyzed and dynamically updated to identify local areas where abrupt concentration gradients still exist, thus providing precise target localization for subsequent reverse vibration intervention. This process specifically includes the following steps:

[0088] Based on the initial concentration distribution matrix constructed in the previous step, the spatial gradient of solute concentration at various locations on the membrane module surface is calculated using numerical gradient analysis. Each grid point in the concentration distribution matrix is ​​treated as a data unit, and the concentration difference between that point and its adjacent units (up, down, left, and right) is used to calculate the local concentration gradient at that point. By setting a concentration gradient threshold (e.g., ≥15 mg / L·cm), sensitive regions with abrupt concentration changes on the membrane surface can be identified. The judgment criteria include not only whether the gradient intensity exceeds the set threshold, but also whether the direction of concentration change points to a fixed center, thereby distinguishing between natural diffusion and anomalous aggregation, ensuring that the identified regions have a physical basis for crystallization.

[0089] Numerical gradient analysis is a method that judges the spatial trend and severity of a physical quantity's change by comparing the differences in data from adjacent measuring points. In this step, this method is used to analyze the solute concentration distribution at various locations on the surface of the membrane module, identifying areas with significant concentration abrupt changes, thereby determining whether there is a risk of crystallization. Its role is to convert abnormal change patterns in the concentration distribution matrix, which are not easily detected directly, into identifiable high-risk indicators through quantitative means, facilitating the precise identification of locations requiring key intervention.

[0090] The specific steps are as follows:

[0091] The surface of the membrane module is divided into a uniform two-dimensional grid, with each grid corresponding to a specific sampling point. The solute concentration data of the sampling point comes from the previous concentration acquisition process.

[0092] The concentration difference between each grid point and its surrounding neighboring points is compared sequentially, and the magnitude of the difference is used to determine whether the point is in a region of drastic concentration change.

[0093] Based on the set concentration change threshold, areas where the change exceeds the preset standard are marked, and these areas are considered to have abnormal concentration accumulation.

[0094] These marked regions are grouped into "abrupt concentration gradient regions" and used as key targets for subsequent reverse vibration interventions. This method is characterized by high computational efficiency, strong adaptability, and ease of integration with real-time data updates, making it particularly suitable for continuous monitoring and assessment of the microscale state of the membrane surface.

[0095] During the ongoing membrane surface turbulence homogenization process, solute concentration data at each measuring point are reacquired at fixed time intervals (e.g., every 30 seconds), and a dynamically updated concentration distribution matrix is ​​constructed using the same method. The new matrix is ​​then compared with the initial matrix to observe the concentration change trends in each region, paying particular attention to whether previously identified abrupt changes show a leveling-off trend or whether they still maintain a high gradient. This method can eliminate false concentration fluctuations caused by short-term perturbations, thereby improving the stability and accuracy of the identification results.

[0096] Regions exhibiting high concentration gradients and a continuously inward-concentrating trend across multiple cycles during dynamic updates are identified as "abnormal concentration residual regions" where the turbulence homogenization effect is insufficient or ineffective. These regions may exhibit microscale stagnation or local solute re-aggregation, making them high-probability locations for crystal nucleation and accumulation. To avoid omissions, the concentration gradient evolution curves of these regions will be fitted to determine whether they exhibit unstable trends such as nonlinear growth or periodic fluctuations, further enhancing the robustness of identification.

[0097] The spatial coordinates, concentration gradient intensity, duration, and evolution characteristics of the identified residual areas of concentration anomalies are integrated into an intervention data package, which serves as a precise target input for subsequent reverse vibration intervention operations. This data is not only used to select the driving position of specific vibration devices, but also, based on the temporal characteristics of the concentration anomaly evolution, to determine the timing and intensity of reverse intervention, achieving a more targeted and efficient asymmetric disturbance response.

[0098] This step aims to continuously identify and dynamically track the solute distribution on the membrane module surface, ensuring that even after spatial focusing perturbation, risk areas with incomplete homogenization of local concentration anomalies can still be effectively captured. This provides precise, real-time, and targeted target location for reverse vibration intervention. Because sodium methoxide exhibits a strong crystallization tendency at high concentrations, even after initial perturbation homogenization, some membrane areas may still exhibit microscale stagnation, solute re-aggregation, or local perturbation blind zones, resulting in a "residual high gradient" phenomenon in space. These areas are highly susceptible to becoming initial nucleation sites for crystals, leading to bridging structures and membrane pore blockage. By dynamically updating the concentration distribution matrix in this step, not only can the trend of solute concentration changes with perturbation intervention be reflected in real time, but numerical gradient analysis can also be used to identify local points where concentration abrupt changes have not been eliminated, thus providing feedback verification of the concentration control effect during perturbation. Furthermore, areas that maintain a high gradient or exhibit abnormal trends across multiple sampling cycles are extracted as target locations for implementing reverse vibration intervention. Compared to traditional methods that rely on static data to develop intervention plans, this step establishes a closed-loop control path of "perturbation-response-correction," making membrane risk identification more continuous and adaptable.

[0099] S5, targeting the abnormal concentration gradient region extracted from the concentration distribution matrix, activates the reverse pulse vibration intervention mechanism, applies short-period high-energy reverse pulse vibration to the target region to destroy the initial crystal nucleus structure in the solute aggregation process in the region, and prevents the formation of crystal bridging structure and the blockage of filter pores.

[0100] To prevent the formation of initial crystal nuclei and the accumulation of crystal bridging structures caused by abnormal solute concentration in localized areas of the membrane module surface due to sodium methoxide solution, which could lead to irreversible blockage of the filter pores, targeted intervention is performed on identified high-risk areas using short-cycle, high-energy reverse pulse vibration. The specific steps include:

[0101] Based on the abnormal concentration gradient regions extracted from the dynamic concentration distribution matrix, the target intervention area is precisely located within the spatial coordinate range of the membrane module surface. The abnormal concentration region is a localized area that still exhibits a high concentration gradient, poor stability, or abnormal trend after multiple rounds of turbulence homogenization. Its concentration increase trend displays a non-diffuse distribution characteristic, consistent with typical physical precursors of initial crystal nucleation. In this step, the intervention priority is further evaluated based on the direction and intensity of concentration changes, providing a basis for pulse energy setting.

[0102] Based on the positioning results, a reverse pulse vibration intervention mechanism is activated. "Reverse pulse vibration" refers to the energy release in the opposite direction to the conventional focusing perturbation direction, creating an impact-type disturbance in the target area through short-duration, high-amplitude transient vibration waves, distinct from normal continuous wave vibration. This vibration is applied to the corresponding region of the membrane surface with an extremely short period (e.g., completing one pulse cycle within 0.5 seconds), driving the liquid phase to generate asymmetric perturbed eddies, disrupting the forming solute aggregation trend, and forcibly destroying the statically stable structural environment required for crystal nucleation.

[0103] During the application of reverse pulse vibration, the instantaneous kinetic energy transfer through high-frequency vibration disrupts the weak van der Waals forces between solute particles, blocking their self-assembly path under critical concentration conditions. In this process, the pulsed pressure wave generated by the vibration propagates along the boundary layer, inducing rapid fluid disturbance and concentration dilution effects within the liquid-phase microregions on the membrane surface. This significantly reduces the initial nucleation probability and inhibits subsequent crystal aggregation and structural bonding, preventing crystal bridging growth between membrane pores and thus avoiding flux blockage.

[0104] After the intervention is completed, concentration monitoring continues in the target area, and an updated concentration distribution matrix is ​​generated to determine whether the concentration gradient in the area has significantly decreased and tended to be uniformly distributed. If the concentration gradient still exists, the next round of reverse pulse intervention is automatically set, or the pulse energy is increased or the intervention time is extended to ensure that the concentration anomaly is completely eliminated.

[0105] This step aims to proactively disrupt the initial nucleation structure of crystals that may form during the accumulation of solute concentration in localized areas on the membrane module surface through targeted physical intervention. This prevents the formation of crystal bridging structures at the source, avoiding serious problems such as filter pore blockage and sudden flux drops. In the dynamic separation and impurity control of sodium methoxide, although preliminary regulation of the membrane surface has been achieved through methods such as turbulence homogenization and concentration distribution matrix identification, some microscale regions may still exhibit characteristics such as high concentration, abrupt gradient changes, and unstable evolution trends under turbulence. These regions are highly susceptible to crystallization and become potential trigger points for failure. This step activates a reverse pulse vibration intervention mechanism, generating strong disturbances locally in abnormal concentration areas with extremely short-cycle, high-energy-density, and directional reverse pulse waves. This disturbance can instantly disrupt the stable environment and spatial structure required for crystal nucleation between solute particles, weakening the adsorption forces and alignment tendencies between particles, thereby effectively delaying or terminating the crystal formation process. Furthermore, the short-duration pressure waves and shear forces induced by pulsed vibration can create a high-speed micro-perturbation flow field near the membrane surface, enhancing the diffusion effect of the liquid phase, accelerating the redistribution of solutes, effectively eliminating local supersaturation, and restoring the thermodynamic stability of the region. Compared with traditional uniform vibration or overall perturbation methods, this step has stronger localization, higher intervention intensity, and faster response speed. It can intervene before the crystals have stabilized, improving the foresight and proactivity of process control. This is a key technical node to ensure the long-term stable operation of the membrane module and maintain the output of high-purity products.

[0106] S6, based on the changing trends of the temperature distribution matrix and concentration distribution matrix after the reverse pulse vibration intervention, jointly analyzes the response characteristics of each high-risk area, dynamically corrects the frequency output parameters and temperature control strategy of the vibration device, and realizes the closed-loop precise control of the entire process of crystal precipitation on the surface of the membrane module to maintain the separation stability and purification efficiency during long-term continuous operation.

[0107] To achieve closed-loop control of crystal precipitation behavior on the membrane module surface during sodium methoxide preparation and ensure high stability and purification efficiency under long-term continuous operation, temperature and concentration data after reverse pulse vibration intervention were integrated to dynamically analyze the intervention response characteristics in high-risk areas and adjust control parameters accordingly. The specific steps include:

[0108] After the reverse pulse vibration intervention was completed, multi-point synchronous temperature and concentration data were collected on the membrane module surface, and the temperature distribution matrix and concentration distribution matrix were updated accordingly. Temperature data collection covered all high-risk areas where pulse vibration was applied. By comparing the temperature change curves before and after the intervention, response characteristics such as recovery from thermal anomalies, enhanced boundary heat conduction, or mitigation of local temperature rises were identified during the intervention. Simultaneously, the concentration change trend at the same location was analyzed to determine whether the concentration tended to be uniform and whether the concentration gradient decreased below the safe threshold. The combined trend of the two matrices can be used to comprehensively evaluate the intervention effect and the degree of recovery stability in the region.

[0109] Based on the aforementioned temperature and concentration change trends, a response characteristic assessment model was established for each high-risk area. This model uses parameters such as response intensity (e.g., temperature fluctuation range, concentration gradient descent rate), post-intervention hysteresis recovery time, residual concentration peak, and fluctuation frequency as variables. By setting multi-dimensional assessment criteria, the intervention effect in each area is classified into three categories: "complete response," "partial response," and "no response." Differences in response characteristics will directly affect the subsequent adjustment direction and intervention intensity of the vibration control strategy, providing a basis for precise feedback control.

[0110] Based on the judgment results, the frequency output parameters of the vibration device are dynamically corrected. In the "full response" region, the vibration energy density is appropriately reduced or the intervention cycle interval is extended to reduce energy consumption and decrease the mechanical load on the membrane surface. In the "partial response" region, the current frequency setting is maintained and the intervention duration is extended. In the "no response" region, the vibration frequency is increased, the pulse energy is increased, or the vibration waveform combination is changed to create a stronger local disturbance field. Simultaneously, combined with the temperature response characteristics, the membrane surface heat conduction path, boundary cooling strategy, or heat transfer power is adjusted to control the temperature field redistribution and achieve coordinated control of heat and mass coupling.

[0111] The aforementioned corrected parameters are input into the control process in real time, and a new round of temperature and concentration data acquisition and matrix reconstruction continues, forming a periodic closed-loop control chain of monitoring-judgment-parameter adjustment-verification. Each adjustment action not only responds to the current state but also predicts future risk evolution trends through time series data trends, allowing for early intervention and adjustment to delay or avoid the formation of the critical state for crystal precipitation to the greatest extent possible.

[0112] This step aims to construct a data-feedback-based intelligent control closed loop, enabling continuous sensing, analysis, and precise control of crystal precipitation behavior on the membrane module surface during dynamic operation. This ensures long-term stable operation of the membrane separation process and continuous guarantee of product purity. In the post-processing stage of sodium methoxide preparation, local crystallization risk often exhibits periodic fluctuations and sudden enhancements. Even with interventions such as reverse pulse vibration on abnormal areas of the membrane surface, it is difficult to guarantee that all areas can immediately return to a stable state. To avoid secondary crystallization caused by residual crystal nuclei or the emergence of new risk areas, continuous monitoring and response analysis of the membrane surface state after reverse intervention are necessary. By collecting the temperature and concentration distribution matrices after vibration intervention and comparing the data differences before and after intervention, it is possible to assess whether high-risk areas have responded as expected to the intervention. Furthermore, by extracting key indicators such as temperature mitigation rate, concentration gradient smoothness, and recovery time, the response effects of different areas are classified and analyzed. The vibration frequency, pulse energy, duration of action, and membrane surface temperature control strategies are then dynamically adjusted based on the response characteristics. This process enables adaptive adjustment of control parameters, eliminating reliance on manual experience or fixed templates. This allows the membrane module to make precise adjustments in response to varying concentration distributions, disturbances, and operating conditions, continuously maintaining thermodynamic stability at the separation interface, effectively extending membrane life and preventing product quality fluctuations. In summary, this step, by establishing a closed-loop control mechanism based on the fusion of physical feedback and intelligent judgment, truly transforms membrane surface crystallization behavior from "passive inhibition" to "active management," which is the core element of this invention for improving process robustness and operational continuity.

[0113] Experimental data illustrate the control of impurities and intervention of crystallization risk in the dynamic membrane separation process of sodium methoxide.

[0114] I. Experimental Objective

[0115] The effectiveness of the temperature distribution matrix and concentration distribution matrix identification and intervention method described in this invention is verified. The control effect of suppressing membrane surface crystal precipitation by spatial focusing perturbation and reverse pulse intervention is evaluated, and key performance indicators (flux recovery rate, blockage rate reduction, membrane lifetime extension, etc.) are quantified.

[0116] II. Experimental Setup and Parameter Setting

[0117] Table 1: Summary Table of Key Equipment and Process Parameters in the Embodiments

[0118]

[0119] III. Overview of Experimental Procedures

[0120] Initial state acquisition: Record the temperature distribution and solute concentration distribution on the surface of the membrane module without any vibration interference;

[0121] Focused disturbance intervention phase: Implement spatial focused disturbances on the identified warming areas;

[0122] Concentration homogenization monitoring phase: Record changes in concentration distribution and form an initial concentration distribution matrix;

[0123] Reverse pulse intervention phase: Apply reverse pulses to areas where high concentration gradients still exist;

[0124] Closed-loop control adjustment phase: Adjust the intervention frequency and energy based on temperature / concentration trend feedback.

[0125] IV. Key Experimental Data Results

[0126] 1. Temperature distribution data as shown in Table 2 below (examples of some areas)

[0127] Table 2:

[0128]

[0129] High-risk area identification criteria: temperature rise ≥ +0.8℃; response criteria: temperature drop >1.5℃ is considered effective.

[0130] 2. The concentration distribution and gradient change data are shown in Table 3 below.

[0131] Table 3:

[0132]

[0133] The gradient threshold was set at 15 mg / L·cm; if it was below this value, homogenization was considered effective.

[0134] 3. The following table 4 shows the throughput data statistics (sampled every 5 minutes).

[0135] Table 4:

[0136]

[0137] Intervention can significantly increase the throughput in the blocked area, avoiding downtime for routine cleaning.

[0138] 4. The crystallinity variation trend is shown in Table 5 below (observed sample surface).

[0139] Table 5:

[0140]

[0141] V. Conclusion

[0142] The results of this experiment show that:

[0143] By deploying multi-point temperature and concentration acquisition devices and constructing a two-dimensional matrix, high-risk areas for film surface crystallization can be clearly identified.

[0144] Spatial focusing perturbation can achieve local concentration homogenization and thermal state regulation, effectively suppressing crystal nucleus formation;

[0145] Reverse pulse vibration significantly intervenes in the residual area of ​​abnormal concentration, and the concentration gradient decreases significantly;

[0146] By combining closed-loop analysis and parameter self-adjustment mechanism, the filtration flux recovery rate is increased by more than 30%, and the membrane service life is extended by an estimated 1.7 times.

[0147] The crystal formation density decreased significantly, effectively controlling the "bridging" blockage phenomenon.

[0148] The impurity control method described above in the sodium methoxide preparation process effectively constructs an integrated regulation mechanism encompassing microscopic state perception, intelligent identification, efficient intervention, and closed-loop control. This mechanism enables precise regulation of the crystallization and precipitation behavior on the membrane module surface throughout the entire process, yielding significant beneficial effects. Specifically, this method, through the real-time construction and updating of temperature and concentration distribution matrices, overcomes the limitations of traditional membrane separation processes that "can only intervene after the fact and cannot dynamically predict." It achieves, for the first time, a high-resolution visualization of the membrane surface thermodynamics and substance concentration state, providing a scientific basis for risk identification. Furthermore, based on the spatial positioning of high-risk areas, a localized precise control method of "spatial focusing perturbation + reverse pulse intervention" is employed. This proactively intervenes before the crystals stabilize, breaking the initial crystal nucleus structure and significantly reducing the risks of crystal bridging and membrane blockage. Simultaneously, through feedback analysis of the intervention results, the vibration frequency and temperature control strategy are dynamically corrected, constructing a closed-loop regulation system of "data-driven—parameter adaptive—process steady-state maintenance." Overall, this invention not only improves the anti-crystallization ability and service life of membrane modules, but also effectively reduces the cleaning frequency and downtime risk while ensuring separation purity and continuous operation, demonstrating outstanding industrial adaptability and application value.

[0149] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for controlling impurities in the preparation process of sodium methoxide, characterized in that, Includes the following steps: S1. Collect temperature data at multiple points on the surface of the membrane module, construct a temperature distribution matrix, and identify high-risk areas with the risk of crystallization precipitation; High-sensitivity micro thermocouple temperature sensors are deployed on the surface of the membrane module to collect multi-point temperature data in real time. A two-dimensional interpolation algorithm is used to generate and visualize the temperature distribution matrix. Based on the conditions that the local temperature is 0.8 degrees Celsius higher than the overall average, or the temperature difference between adjacent areas is greater than 5 degrees Celsius per centimeter, or the temperature rises continuously for more than 2 minutes without a downward inflection point, high-risk areas for crystallization are identified and their coordinates and temperature change trends are recorded. S2, for the identified high-risk areas, adjust the local frequency output of the vibration device to generate directional spatial focusing disturbance in the target area, and realize local temperature disturbance intervention; S3 induces microscale turbulence in the liquid phase under the effect of spatial focusing disturbance, dynamically homogenizes the solute concentration in the target area, and simultaneously collects concentration data to construct an initial concentration distribution matrix; S4. Based on the initial concentration distribution matrix, identify regions where abrupt concentration gradients still exist, and dynamically update the concentration distribution matrix during the turbulence averaging process to extract abnormal concentration regions and determine the location of reverse intervention. S5 applies short-cycle, high-energy reverse pulse vibrations to the abnormal concentration region of the extract, thereby disrupting the initial crystal nucleus structure and inhibiting crystal bridging and membrane pore blockage. S6, based on the trend of temperature and concentration matrix changes after reverse intervention, jointly analyzes the response of each high-risk area, dynamically corrects the vibration frequency and temperature control strategy, and achieves closed-loop precise control of the crystal precipitation process.

2. The impurity control method in the sodium methoxide preparation process according to claim 1, characterized in that, The specific steps for using a two-dimensional interpolation algorithm to compensate for the temperature between sampling points and generate a complete temperature distribution matrix are as follows: The surface of the membrane module is established as a two-dimensional coordinate plane, and each temperature acquisition point is mapped to a two-dimensional coordinate point with a fixed position and assigned a corresponding temperature value. Choose one of the interpolation methods, such as bilinear interpolation, bicubic interpolation, or Gaussian kernel regression interpolation, and calculate the temperature value at the unknown location based on the spatial relationship between known points to achieve a smooth transition and continuous distribution of the temperature field. Based on the actual size of the membrane module, the membrane surface is divided into a fine grid, and the temperature value of each grid point is calculated using an interpolation function to construct a temperature distribution matrix containing temperature data from all locations.

3. The impurity control method in the sodium methoxide preparation process according to claim 1, characterized in that, Step S2 includes: Two-dimensional coordinates and temperature change trends are extracted from high-risk areas identified in the temperature distribution matrix. Input the coordinates of the high-risk area into the vibration control unit, call the pre-established vibration frequency response model, and calculate the optimal frequency output value and vibration energy parameters; The driving vibration device generates spatial focusing disturbance above the target area, which acts on the liquid boundary layer to disrupt local thermal accumulation and concentration retention. Monitor the temperature change trend of the target area. If the temperature tends to stabilize, maintain the current frequency output; otherwise, recalculate and adjust the frequency parameters.

4. The impurity control method in the sodium methoxide preparation process according to claim 1, characterized in that, Step S3 includes: Based on the location of high-risk areas, the vibration device outputs spatial focusing disturbances with wavelengths of 1 to 10 millimeters and amplitudes of 10 to 500 micrometers to form a microscale turbulent flow field in the liquid phase boundary layer of the membrane surface; During the turbulence process, high-precision concentration sensors are used to simultaneously collect solute concentration data at multiple locations in the target area and perform normalization processing. The collected concentration data is mapped to the two-dimensional coordinate plane of the membrane module, and a two-dimensional interpolation algorithm is used to estimate the concentration values ​​between sampling points; An initial concentration distribution matrix is ​​generated according to the membrane surface grid division rules to reflect the solute distribution state of the liquid phase on the membrane module surface.

5. The impurity control method in the sodium methoxide preparation process according to claim 1, characterized in that, Step S4 includes: The initial concentration distribution matrix is ​​divided into a two-dimensional grid, and the concentration gradient between each grid point and its adjacent points is calculated. Based on the set concentration gradient threshold, identify abrupt change regions and determine whether there is a concentration change trend that concentrates inward. During the turbulence homogenization process, concentration data is collected at fixed intervals and the concentration distribution matrix is ​​dynamically updated. Regions that maintain a high concentration gradient over multiple consecutive cycles are selected as target locations for reverse vibration intervention.

6. The impurity control method in the sodium methoxide preparation process according to claim 1, characterized in that, Step S5 includes: Abnormal concentration gradient regions are located and intervention priorities are set based on dynamically updated concentration distribution matrices; Short-period, high-energy reverse pulse vibrations are applied to the target region to drive the liquid phase to generate asymmetric perturbation eddies. By using pulsed kinetic energy to disrupt the aggregated structure of solute particles, the crystal nucleation path is blocked and crystal bridging is prevented. After intervention, continue to collect concentration data and determine the trend of concentration gradient changes to decide whether to carry out subsequent intervention operations.

7. The impurity control method in the sodium methoxide preparation process according to claim 1, characterized in that, Step S6 includes: Collect temperature and concentration data after reverse pulse vibration intervention and update the temperature distribution matrix and concentration distribution matrix; Based on the trend of matrix changes, a response characteristic judgment model is established and the intervention effect levels are classified; The frequency output parameters and temperature control strategy of the vibration device are dynamically adjusted according to the response level. The corrected parameters will be input into the control process, and a new round of monitoring, judgment, and control operations will be executed.

Images