A method for preparing mercaptoacetic acid by two-stage acidification double extraction of sodium mercaptoacetate and hydrogen chloride

Abstract

The application discloses a method for preparing mercaptoacetic acid by two-stage acidification and double extraction of sodium mercaptoacetate and hydrogen chloride, relates to the technical field of mercaptoacetic acid preparation, and comprises raw material pretreatment, in-situ microchannel two-stage acidification and synergistic antioxidant reaction, ultrasonic and magnetic nanometer air floatation impurity removal, amphiphilic ionic liquid composite double extraction, ultrasonic and gradient vacuum stripping distillation, MIPs adsorption and short-path molecular distillation refining, membrane distillation and electrodialysis salt recovery and a digital twin intelligent control system.In the application, precise temperature control acidification of a microchannel is cooperated with a composite antioxidant system to inhibit oxidative decomposition from the source and improve product stability; meanwhile, multi-stage impurity removal is combined with composite double extraction to solve the problem of impurity interference, greatly improve the recovery rate and purity of the target product; then, gradient vacuum stripping distillation is combined with ultrasonic mass transfer enhancement to realize low-temperature solvent recovery, reduce energy consumption and solvent loss; and finally, membrane distillation and electrodialysis processes are used to realize wastewater resource utilization and reduce environmental protection treatment cost.

Description

Technical Field

[0001] This invention relates to the field of thioglycolic acid preparation technology, specifically a method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride through a two-stage acidification and double extraction process. Background Technology

[0002] The sodium thioacetate hydrochloric acidification method is currently the mainstream method for industrial production of thioacetic acid. The process mainly involves using sodium thioacetate aqueous solution as raw material, undergoing initial hydrochloric acid acidification, air flotation for impurity removal, and secondary hydrochloric acid acidification to obtain a mixed solution of thioacetic acid and brine. The target product is separated by solvent pretreatment extraction and secondary extraction. The extract is purified by distillation, back-extraction distillation, vacuum dehydration, and thin-film evaporation to obtain the thioacetic acid product. The secondary product is returned to the acidification cycle, the saline wastewater is sent to the sewage treatment plant, the solvent is partially recovered, and the residue at the bottom of the reactor is neutralized.

[0003] However, in existing technologies, the sulfur atoms in the thiol group of the mercaptoacetic acid molecule have high electron cloud density and low valence state, making them prone to oxidation reactions due to electron loss. Furthermore, the carbon chain connecting the thiol group and the carboxyl group is short, and the chemical bond is unstable, making it susceptible to breakage and decomposition at high temperatures due to increased molecular kinetic energy. Current preparation methods have not developed a suitable technical system to address this specific physicochemical property. Instead, they employ generalized processes such as batch acidification, traditional flocculation and flotation, single solvent extraction, and basic temperature and pressure controlled distillation. This mismatch between generalized processes and specific requirements directly leads to uneven mixing during batch acidification, causing localized fluctuations in reaction conditions and incomplete removal of macromolecular impurities. Coupled with the insufficient selectivity of single solvent extraction, this not only interferes with the separation of the target product but also further exacerbates the oxidation risk and causes product decomposition. Ultimately, this not only results in a prominent problem of oxidative decomposition of thiolacetic acid, making it difficult to improve the recovery rate and purity of the target product, but also triggers a chain reaction of high solvent consumption and energy consumption. Summary of the Invention

[0004] The purpose of this invention is to provide a method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via a two-stage acidification and double extraction process, in order to solve the problems in the prior art mentioned in the background section.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via a two-stage acidification and double extraction process, comprising the following steps: S1. Raw material pretreatment: Filter a 20% sodium thioacetate aqueous solution through a 5μm precision filter membrane; S2, In-situ microchannel two-stage acidification and synergistic antioxidant reaction: The pretreated sodium thioacetate aqueous solution and 31% hydrochloric acid were introduced into the microchannel reactor. The pH of the first acidification was adjusted to 3.2±0.1 and the pH of the second acidification was adjusted to 1.5±0.1 by an online Raman spectroscopy pH monitor. A composite antioxidant was added. S3. Ultrasonic and magnetic nano-air flotation for impurity removal: After ultrasonic pretreatment, the acidified liquid is mixed with silica-coated iron oxide magnetic nano-flocculators and separated by an external magnetic field. S4. Amphiphilic ionic liquid composite dual extraction: Using [BMIM][BF4] ionic liquid and MIBK composite extractant, two-stage extraction is performed through a dynamic ceramic membrane extractor; S5. Ultrasonic and gradient vacuum back-extraction distillation: The extract is subjected to gradient vacuum control in an ultrasonic coupled distillation column to recover the solvent and obtain a crude mercaptoacetic acid aqueous solution; S6, MIPs adsorption and short-path molecular distillation purification: Crude mercaptoacetic acid is adsorbed by TGA molecularly imprinted polymer and then purified by short-path molecular distillation. S7. Membrane distillation and electrodialysis salt recovery: saline wastewater is concentrated by membrane distillation and purified by electrodialysis to recover food-grade salt and reuse pure water; S8, Digital Twin Intelligent Control: Optimizes the entire process operation parameters in real time through a digital twin model.

[0006] Preferably, the microchannel reactor has a channel size of 50-100 μm, a hydrochloric acid dropping rate of 10-15 L / h for the first acidification, a hydrochloric acid dropping rate of 8-12 L / h for the second acidification, and a reaction temperature of 20-25 °C.

[0007] Preferably, the composite antioxidant is composed of sodium L-cysteine, sodium bisulfite, and EDTA-2Na in a mass ratio of 2:1:0.5, and the amount added is 0.05-0.1% of the total mass of the acidification solution.

[0008] Preferably, the ultrasonic pretreatment parameters are 18-22kHz, 250-350W, and 8-12min, the silica-coated iron oxide magnetic nano-flocculator has a particle size of 40-60nm, the addition amount is 50-80mg / L, the external magnetic field strength is 0.25-0.35T, and the separation time is 6-10min.

[0009] Preferably, the volume ratio of [BMIM][BF4] ionic liquid to MIBK in the composite extractant is 1:4, the volume ratio of pretreatment extraction solvent to feed liquid is 1:2-1:3, the volume ratio of secondary extraction solvent to raffinate is 1:1-1:2, the dynamic ceramic membrane extractor has a membrane pore size of 200nm and a membrane surface flow rate of 0.5m / s.

[0010] Preferably, the ultrasonic parameters of the ultrasonic coupled distillation column are 25-31 kHz and 450-550 W; the pre-extraction distillation vacuum degree is -0.075 to -0.085 MPa and the temperature is 63-72℃; the back-extraction distillation vacuum degree is -0.090 to -0.100 MPa and the temperature is 53-62℃.

[0011] Preferably, the TGA molecularly imprinted polymer uses mercaptoacetic acid as the template molecule, ethylene glycol dimethacrylate as the crosslinking agent, adsorption temperature is 25-30℃, space velocity is 1-2 hL / (L・h), short-path molecular distillation membrane scraping speed is 1400-1600 r / min, vacuum degree is -0.09 to -0.10 MPa, and heating temperature is 60-65℃.

[0012] Preferably, the membrane distillation unit uses a PVDF hollow fiber membrane with a pore size of 0.15-0.25 μm, an operating temperature of 65-75℃, and a vacuum degree of -0.085 to -0.095 MPa; the electrodialysis unit has an ion exchange capacity of 1.8-2.2 mmol / g.

[0013] Preferably, the digital twin model is built based on Python, integrating reaction kinetics and mass and heat transfer equations, and collects data in real time through a TGA concentration sensor, a -SH group content sensor, and an impurity ion sensor, with a data acquisition interval of 5 seconds; The calculation formula for the digital twin model is as follows: ; In the formula: The total material change rate of TGA within the system, in mol / s; The total mass of materials within the process system, in kg; The mole fraction of TGA; Time, in seconds; The effective volume of the acidification reaction unit is expressed in cubic meters (m³). 3 ; The rate constant for the acidification reaction is denoted as . Unit m 3 / (mol·s); This represents the molar concentration of sodium thioglycolate, in mol / m³. 3 ; This refers to the molar concentration of hydrochloric acid, in mol / m³. 3 ; The mass transfer coefficient of the extraction / membrane distillation unit is expressed in m / s. The effective mass transfer area is expressed in meters (m²). 2 △cTGA represents the TGA concentration difference during the mass transfer process, in mol / m³. 3 ; The thermal decomposition coefficient of TGA; R represents the total heat transfer of the distillation unit, in W; R is the gas constant, taken as 8.314 J / (mol·K). The operating temperature of the process unit is expressed in Kelvin (K). This refers to the loss of TGA mole fraction due to thermal decomposition. To adjust the correction coefficient, the digital twin model output is set to 0.8-1.2; The increment of operating parameters resulting from intelligent regulation, in mol / s.

[0014] Preferably, the digital twin intelligent control system includes a physical entity layer, a sensor network layer, an edge computing and data preprocessing layer, a digital twin model core layer, an optimization decision and control layer, an actuator and command instruction layer, a data storage and historical database, and system functions and performance indicators. The physical entity layer serves as the physical carrier for the operation of the entire process unit and provides controllable objects. The sensor network layer, based on TGA concentration, -SH group content, and impurity ion sensors, collects core parameters at 5-second intervals and transmits them to the edge computing and data preprocessing layer. The edge computing and data preprocessing layer cleans, completes, and standardizes the raw data before synchronously transmitting it to the digital twin model core layer and the data storage and historical database. The digital twin model core layer, built using Python, integrates reaction kinetics and mass and heat transfer equations, calls the preprocessed data, substitutes it into the core calculation formula to complete numerical solutions and process deviation analysis. The results are transmitted to the optimization decision and control layer. Based on the deviation, the optimization decision and control layer outputs the incremental operation parameter ΔUctrl and adjusts the control correction coefficient γ (0.8-1.2), which is then converted into control commands and transmitted to the actuator and command layer. The actuator and command layer drive the actuator to adjust the process parameters of the physical entity layer, and at the same time, the execution results are fed back to trigger the sensor network layer to re-collect data to form a closed loop. The data storage and historical database store the entire process data and provide data support and traceability. The system functions and performance indicators limit the core function boundaries and data acquisition interval ≤5 seconds and model calculation deviation ≤±3% performance thresholds.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. In this invention, precise temperature-controlled acidification through microchannels and a composite antioxidant system work synergistically to inhibit oxidative decomposition at the source and improve product stability. At the same time, multi-stage impurity removal and composite dual extraction are combined to solve the problem of impurity interference, significantly improving the recovery rate and purity of the target product. Then, gradient vacuum back-extraction distillation combined with ultrasonic enhanced mass transfer achieves low-temperature solvent recovery, reducing energy consumption and solvent loss. Membrane distillation and electrodialysis processes are used to realize wastewater resource utilization, reducing environmental treatment costs. Furthermore, digital twin closed-loop control ensures optimal parameters throughout the process, improving process stability and product quality consistency, and completely breaking through the bottleneck of insufficient adaptability of general processes. 2. In this invention, the oxidative and thermosensitive properties of mercaptoacetic acid are precisely adapted. Through high-frequency data acquisition and rapid control response, oxidative decomposition caused by process fluctuations is suppressed in a timely manner, significantly improving product quality stability and recovery rate. Relying on data preprocessing and precise model calculation, process deviations are accurately identified, avoiding the misoperation of traditional experience-based control, ensuring optimal parameters throughout the process, and improving process stability. Supported by full-process data traceability and model iterative optimization, continuous process improvement is facilitated, reducing production trial-and-error costs. Furthermore, solvent loss and energy consumption are reduced through closed-loop control, and resource reuse in the salt recovery unit improves process economy. This breakthrough overcomes the bottleneck of insufficient adaptability of traditional general processes and provides intelligent technical support for the efficient preparation of TGA. Attached Figure Description

[0016] Fig. 1 This is a flowchart of a method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to the present invention; Fig. 2 This is a schematic diagram of the digital twin intelligent control system in the method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to the present invention; Fig. 3 This is a schematic diagram of the digital twin intelligent control system in the method for preparing thioglycolic acid from sodium thioglycolic acid and hydrogen chloride via two-stage acidification and double extraction according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1: Refer to Figs. 1-3 The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via a two-stage acidification and double extraction process includes the following steps: S1. Raw material pretreatment: Filter a 20% sodium thioacetate aqueous solution through a 5μm precision filter membrane; S2, In-situ microchannel two-stage acidification and synergistic antioxidant reaction: The pretreated sodium thioacetate aqueous solution and 31% hydrochloric acid were introduced into the microchannel reactor. The pH of the first acidification was adjusted to 3.2±0.1 and the pH of the second acidification was adjusted to 1.5±0.1 by an online Raman spectroscopy pH monitor. A composite antioxidant was added. The microchannel reactor has a channel size of 50-100 μm. The initial acidification hydrochloric acid dropping rate is 10-15 L / h, and the secondary acidification hydrochloric acid dropping rate is 8-12 L / h. The reaction temperature is 20-25℃. The composite antioxidant is composed of sodium L-cysteine, sodium bisulfite, and EDTA-2Na in a mass ratio of 2:1:0.5, and the addition amount is 0.05-0.1% of the total mass of the acidification solution. S3. Ultrasonic and magnetic nano-air flotation for impurity removal: After ultrasonic pretreatment, the acidified liquid is mixed with silica-coated iron oxide magnetic nano-flocculators and separated by an external magnetic field. The ultrasonic pretreatment parameters are 18-22kHz, 250-350W, and time of 8-12min. The particle size of the silica-coated iron oxide magnetic nano-flocculant is 40-60nm, the addition amount is 50-80mg / L, the external magnetic field strength is 0.25-0.35T, and the separation time is 6-10min. S4. Amphiphilic ionic liquid composite dual extraction: Using [BMIM][BF4] ionic liquid and MIBK composite extractant, two-stage extraction is performed through a dynamic ceramic membrane extractor; The volume ratio of [BMIM][BF4] ionic liquid to MIBK in the composite extractant is 1:4, the volume ratio of pretreatment extraction solvent to feed liquid is 1:2-1:3, the volume ratio of secondary extraction solvent to raffinate is 1:1-1:2, the dynamic ceramic membrane extractor has a membrane pore size of 200nm and a membrane surface flow rate of 0.5m / s. S5. Ultrasonic and gradient vacuum back-extraction distillation: The extract is subjected to gradient vacuum control in an ultrasonic coupled distillation column to recover the solvent and obtain a crude mercaptoacetic acid aqueous solution; The ultrasonic parameters of the ultrasonic coupled distillation column are 25-31 kHz and 450-550 W; the vacuum degree of pre-extraction distillation is -0.075 to -0.085 MPa and the temperature is 63-72℃; the vacuum degree of back-extraction distillation is -0.090 to -0.100 MPa and the temperature is 53-62℃. S6, MIPs adsorption and short-path molecular distillation purification: Crude mercaptoacetic acid is adsorbed by TGA molecularly imprinted polymer and then purified by short-path molecular distillation (TGA refers to mercaptoacetic acid). The TGA molecularly imprinted polymer uses mercaptoacetic acid as the template molecule, ethylene glycol dimethacrylate as the crosslinking agent, adsorption temperature 25-30℃, space velocity 1-2 hL / (L・h), short-path molecular distillation membrane scraping speed 1400-1600 r / min, vacuum degree -0.09 to -0.10 MPa, and heating temperature 60-65℃. S7. Membrane distillation and electrodialysis salt recovery: saline wastewater is concentrated by membrane distillation and purified by electrodialysis to recover food-grade salt and reuse pure water; The membrane distillation unit uses PVDF hollow fiber membranes with a pore size of 0.15-0.25 μm, an operating temperature of 65-75℃, and a vacuum degree of -0.085 to -0.095 MPa; the electrodialysis unit has an ion exchange capacity of 1.8-2.2 mmol / g. S8, Digital Twin Intelligent Control System: Optimizes the entire process operation parameters in real time through a digital twin model.

[0019] The working principle and usage of this process: In the raw material pretreatment stage, a 20% sodium mercaptoacetate aqueous solution is filtered using a 5μm precision filter membrane. The core function is to remove large particulate mechanical impurities and suspended matter entrained in the raw materials, avoid clogging of the subsequent microchannel reactor, and reduce the interference of impurities on the acidification reaction. By leveraging the high specific surface area, high mixing efficiency, and rapid mass and heat transfer characteristics of microchannel reactors, the problem of localized pH fluctuations caused by uneven mixing in traditional batch acidification processes is solved. Real-time feedback from an online Raman spectroscopy pH monitor allows for precise control of the initial acidification pH (3.2±0.1, weakly acidic environment for initial conversion of sodium thioglycolate, avoiding direct acidification in strong acidic environments that could lead to excessive localized reactions) and the secondary acidification pH (1.5±0.1, strong acidic environment for complete deep conversion), ensuring the uniform and complete conversion of sodium thioglycolate to thioacetic acid. Simultaneously, the reaction temperature is controlled at 20-25℃ (low temperature inhibits thiol oxidation), and a composite antioxidant of sodium L-cysteine, sodium bisulfite, and EDTA-2Na is used to suppress oxidative decomposition during the acidification process from both temperature control and chemical protection perspectives.

[0020] Ultrasonic vibration at 18-22kHz and 250-350W disperses the agglomerated fine impurity particles in the acidification solution, while simultaneously disrupting the hydration film on the impurity surface, thus improving the binding efficiency between the subsequent flocculant and the impurities. The subsequently added silica-coated iron oxide magnetic nano-flocculator utilizes its high specific surface area and strong adsorption properties to rapidly adsorb large molecular organic impurities and metal ion impurities in the acidification solution. Then, rapid separation is achieved through an external magnetic field of 0.25-0.35T, solving the problems of incomplete impurity removal and low separation efficiency in traditional flocculation and flotation processes, and reducing the interference of impurities on the subsequent extraction process.

[0021] Two-stage extraction is achieved through a dynamic ceramic membrane extractor. The first stage of extraction (solvent to feed volume ratio of 1:2-1:3) achieves the initial enrichment of the target product, while the second stage of extraction (solvent to raffinate volume ratio of 1:1-2) further recovers trace amounts of mercaptoacetic acid in the raffinate, maximizing the recovery rate of the target product.

[0022] The core objective of the back-extraction distillation stage is to recover the solvent and separate the target product at low temperatures, avoiding the decomposition of thioglycolic acid caused by high temperatures. First, the ultrasonic action of the ultrasonic-coupled distillation column can break the gas film formed during distillation, enhance mass transfer efficiency, and reduce the distillation temperature requirement. Then, through pre-extraction distillation, the lower boiling point MIBK solvent is preferentially recovered. Back-extraction distillation realizes the recovery of [BMIM][BF4] ionic liquid and the separation of crude thioglycolic acid aqueous solution. Gradient control ensures efficient solvent recovery (reducing loss) while avoiding the decomposition of thioglycolic acid caused by single high-temperature distillation.

[0023] Using TGA molecularly imprinted polymers (MIPs) with mercaptoacetic acid as a template molecule, specific pores are formed on the surface that match the spatial structure and functional groups of mercaptoacetic acid molecules. This allows for the precise adsorption of trace impurities without adsorbing the target product, thus achieving selective impurity removal. For saline wastewater generated during acidification and extraction processes, a combination of membrane distillation and electrodialysis is used to achieve resource recovery. The vapor pressure difference across the membrane is used to separate water and salt. The operating temperature of 65-75℃ and the vacuum degree of -0.085 to -0.095MPa can efficiently concentrate saline wastewater while avoiding salt crystallization clogging the membrane pores. The electrodialysis unit utilizes the selective permeability of the ion exchange membrane to achieve the directional migration and purification of salt ions under the action of an electric field, realizing the resource utilization of wastewater and reducing environmental treatment costs.

[0024] By combining precise temperature-controlled acidification through microchannels with a composite antioxidant system, oxidative decomposition is inhibited at the source, improving product stability. Simultaneously, multi-stage impurity removal and composite dual extraction are combined to solve the problem of impurity interference, significantly improving the recovery rate and purity of the target product. Then, gradient vacuum back-extraction distillation combined with ultrasonic-enhanced mass transfer achieves low-temperature solvent recovery, reducing energy consumption and solvent loss. Membrane distillation and electrodialysis processes are used to realize wastewater resource utilization, reducing environmental treatment costs. Furthermore, digital twin closed-loop control ensures optimal parameters throughout the process, improving process stability and product quality consistency, completely overcoming the bottleneck of insufficient adaptability of general processes.

[0025] Example 2: Figs. 1-3 As shown, the digital twin intelligent control system optimizes the entire process operation parameters in real time through a digital twin model. The digital twin model is built based on Python and integrates reaction kinetics and mass and heat transfer equations. Data is collected in real time through TGA concentration sensors, -SH group content sensors, and impurity ion sensors, with a data acquisition interval of 5 seconds. The calculation formula for the digital twin model is as follows: ; In the formula: The total material change rate of TGA (thioglycolic acid) in the system is expressed in mol / s. The total mass of materials within the process system, in kg; The mole fraction of TGA; Time, in seconds; The effective volume of the acidification reaction unit is expressed in cubic meters (m³). 3 ; The rate constant for the acidification reaction is denoted as . Unit m 3 / (mol·s); This represents the molar concentration of sodium thioglycolate, in mol / m³. 3 ; This refers to the molar concentration of hydrochloric acid, in mol / m³. 3 ; The mass transfer coefficient of the extraction / membrane distillation unit is expressed in m / s. The effective mass transfer area is expressed in meters (m²). 2 △cTGA represents the TGA concentration difference during the mass transfer process, in mol / m³. 3 ; The thermal decomposition coefficient of TGA; R represents the total heat transfer of the distillation unit, in W; R is the gas constant, taken as 8.314 J / (mol·K). The operating temperature of the process unit is expressed in Kelvin (K). This refers to the loss of TGA mole fraction due to thermal decomposition. To adjust the correction coefficient, the digital twin model output is set to 0.8-1.2; The increment of operating parameters resulting from intelligent regulation, in mol / s.

[0026] The digital twin intelligent control system comprises a physical entity layer, a sensor network layer, an edge computing and data preprocessing layer, a digital twin model core layer, an optimization decision and control layer, an actuator and command layer, a data storage and historical database, and system functions and performance indicators. The physical entity layer serves as the physical carrier for the operation of the entire process unit and provides the control objects. The sensor network layer collects core parameters at 5-second intervals based on TGA (thioglycolic acid) concentration, -SH group content, and impurity ion sensors and transmits them to the edge computing and data preprocessing layer. The edge computing and data preprocessing layer cleans, completes, and standardizes the raw data before synchronously transmitting it to the digital twin model core layer and the data storage and historical database. The digital twin model core layer is based on Python. The system constructs and integrates reaction kinetics and mass and heat transfer equations. It calls preprocessed data and substitutes it into the core calculation formula to complete numerical solutions and process deviation analysis. The results are transmitted to the optimization decision and control layer. The optimization decision and control layer outputs the incremental operation parameter ΔUctrl based on the deviation and adjusts the control correction coefficient γ (0.8-1.2), which is then converted into control commands and transmitted to the actuator and command layer. The actuator and command layer drive the actuator to adjust the process parameters of the physical entity layer. At the same time, the execution results are fed back to trigger the sensor network layer to re-collect data to form a closed loop. The data storage and historical database store the entire process data and provide data support and traceability. The system functions and performance indicators limit the core function boundaries and data acquisition interval ≤5 seconds and model calculation deviation ≤±3% performance thresholds.

[0027] The working principle and usage of this solution: This digital twin intelligent control system is a closed-loop control system adapted to the easily oxidized, thermosensitive and specific physicochemical properties of thioglycolic acid (TGA). Its core positioning is to achieve real-time optimization of the process parameters of the entire TGA preparation process through bidirectional mapping and dynamic interaction between physical entities and virtual models, so as to ensure product quality stability and process efficiency. Its overall logic revolves around the closed-loop link of "sensing-processing-computation-decision-execution-feedback", and the modules work together to form a complete control system.

[0028] The system uses the physical entity layer as its foundation, supporting the operation of the entire TGA preparation process units (raw material pretreatment, two-stage acidification, air flotation for impurity removal, dual extraction, back distillation, purification, and salt recovery). It provides the system with a real process status data source and serves as the final execution object of control commands, realizing the transformation from virtual decision-making to physical operation. The sensor network layer acts as a "sensory bridge," constructing a full-process sensing network based on TGA concentration sensors, -SH group content sensors (using electrochemical thiol sensors, composed of gold nanoparticle-modified working electrodes, reference electrodes, counter electrodes, and signal acquisition modules, adapted to the complex water quality environment of high-salt and high-organic-content mercaptoacetic acid production wastewater), and impurity ion sensors. Equipment is deployed at key quality control nodes such as acidification outlet and extraction outlet, and core parameters are collected in real time at fixed 5-second intervals—TGA concentration directly reflects the formation and loss state of the target product, -SH group content indirectly characterizes the degree of oxidation, and impurity ion concentration reflects the impurity removal effect. The raw electrical signals are transmitted to the edge computing and data preprocessing layer through an industrial bus.

[0029] The edge computing and data preprocessing layer serves as a "data purification hub." After receiving the raw data, it removes outliers using the 3σ criterion, fills in missing values ​​using linear interpolation, and normalizes parameters of different dimensions to the [0,1] interval to eliminate dimensional interference. The high-quality preprocessed data is then synchronously sent to the core layer of the digital twin model for real-time computation and stored in the data storage and historical database to support model iteration and process traceability.

[0030] The core layer of the digital twin model is the system's core. It's built on a Python framework and integrates reaction kinetics and mass and heat transfer equations. The calculation formulas of the digital twin model are the computational core; by substituting preprocessed data into the formulas and combining them with the reaction kinetic sub-equations, the acidification reaction rate constant is calculated. The model then uses the `odeint` function from the Python SciPy library to numerically calculate the total material change rate of the TGA process. This rate directly reflects the overall equilibrium state of TGA generation, mass transfer, and thermal decomposition. The model compares the calculated results with a preset optimal threshold of 0.05-0.1 mol / s to accurately identify deviation types and amounts such as "aggravated oxidation, insufficient acidification, and impurity interference," and transmits this information synchronously to the optimization decision-making and control layer. The optimization decision-making and control layer, acting as the "instruction generation hub," calls the control rule library based on the deviation analysis results and outputs the corresponding operational parameter increments. If oxidation intensifies, the system will output a control command to "reduce distillation temperature and increase antioxidant dosage," while dynamically adjusting the control correction coefficient within the range of 0.8-1.2 based on the magnitude of the deviation. The greater the deviation The closer the value is to 1.2, the stronger the regulation will be, and the smaller the deviation will be. To avoid excessive fluctuations, the control parameters are then converted into industrial control instructions recognizable by the PLC / DCS and transmitted to the actuator and command layer. The actuator and command layer, acting as the "landing carrier," receive the instructions and drive corresponding actuators such as the hydrochloric acid feed regulating valve, antioxidant addition pump, and distillation temperature and pressure controller to adjust process parameters and provide real-time feedback on the action results. This triggers the sensor network layer to re-collect data, completing the closed-loop connection. Data storage and historical databases utilize a time-series database, classifying and retaining full-process data (raw data, preprocessed data, model calculation data, control instructions, etc.) by production batch. This provides historical data support for the model to optimize initial parameters and enables full-cycle traceability of production. System functions and performance indicators define the boundaries of core functions (real-time monitoring, simulation diagnosis, precise control, data traceability) and quantification thresholds (acquisition interval ≤ 5 seconds, calculation deviation ≤ ±3%, response time ≤ 10 seconds) to ensure that the operating effect of each module adapts to process requirements. The system must be used in accordance with the standard process of "initialization-acquisition-calculation-decision-execution-feedback". First, start the physical entity layer equipment to complete the no-load debugging, calibrate the sensors and initialize the model parameters (R=8.314J / (mol·K)). The preset thresholds and performance indicators are set to 0.02, etc.; then the data acquisition and preprocessing program is started, and the data is transmitted and purified at 5-second intervals; the core layer of the model automatically substitutes the data to solve the TGA material change rate and compares the threshold to identify the deviation; the optimization decision layer generates control instructions and converts them into industrial signals.

[0031] After the actuator completes the parameter adjustment, it feeds back the results. The sensor re-collects data to verify the control effect. If the deviation is not eliminated, it iterates and optimizes until it returns to the optimal range. The data storage module synchronously retains the data of the entire process and supports querying and tracing by batch number and iterating the model parameters. The system is tightly logically interconnected: data flows unidirectionally from "physical entity → sensor → edge computing → model → decision → actuator → physical entity," while data is synchronously retained throughout the entire process, forming a "unidirectional control + bidirectional data retention" architecture. Each module is interdependent, with the physical entity layer providing fundamental support. Sensor acquisition accuracy and data preprocessing quality directly affect model calculation accuracy; model deviation identification results determine the rationality of control commands; actuator action accuracy ensures control effectiveness; and data storage and performance indicators guarantee stable system operation. The core formula is the key link connecting the entire process, deeply coupling process unit parameters, sensor data, and control parameters to achieve multi-module collaborative linkage. It can promptly suppress TGA oxidation and decomposition caused by process fluctuations, ensuring that parameters always adapt to their characteristic requirements. Simultaneously, there are collaborative constraints among module parameters, such as the acquisition interval needing to match model calculation time and response time; model calculation deviation and sensor accuracy work together to avoid mis-control, comprehensively ensuring stable and efficient system operation. This helps overcome the bottleneck of insufficient adaptability of traditional general processes, improving product quality and process economy in TGA (thioglycolic acid) preparation.

[0032] Precisely tailored to the easily oxidized and thermosensitive properties of mercaptoacetic acid (HGA), this technology utilizes high-frequency data acquisition and rapid control response to promptly suppress oxidative decomposition caused by process fluctuations, significantly improving product quality stability and recovery rate. Relying on data preprocessing and precise model calculations, it achieves accurate identification of process deviations, avoiding errors from traditional experience-based control, ensuring optimal parameters throughout the process, and enhancing process stability. Full-process data traceability and iterative model optimization support continuous process improvement, reducing production trial-and-error costs. Furthermore, closed-loop control reduces solvent loss and energy consumption, and resource reuse through the salt recovery unit improves process economics. This overcomes the bottleneck of insufficient adaptability of traditional general-purpose processes, providing intelligent technical support for the efficient preparation of TGA.

[0033] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via a two-stage acidification and double extraction process, characterized in that, Includes the following steps: S1. Raw material pretreatment: Filter a 20% sodium thioacetate aqueous solution through a 5μm precision filter membrane; S2, In-situ microchannel two-stage acidification and synergistic antioxidant reaction: The pretreated sodium thioacetate aqueous solution and 31% hydrochloric acid were introduced into the microchannel reactor. The pH of the first acidification was adjusted to 3.2±0.1 and the pH of the second acidification was adjusted to 1.5±0.1 by an online Raman spectroscopy pH monitor. A composite antioxidant was added. S3. Ultrasonic and magnetic nano-air flotation for impurity removal: After ultrasonic pretreatment, the acidified liquid is mixed with silica-coated iron oxide magnetic nano-flocculators and separated by an external magnetic field. S4. Amphiphilic ionic liquid composite dual extraction: Using [BMIM][BF4] ionic liquid and MIBK composite extractant, two-stage extraction is performed through a dynamic ceramic membrane extractor; S5. Ultrasonic and gradient vacuum back-extraction distillation: The extract is subjected to gradient vacuum control in an ultrasonic coupled distillation column to recover the solvent and obtain a crude mercaptoacetic acid aqueous solution; S6, MIPs adsorption and short-path molecular distillation purification: Crude mercaptoacetic acid is adsorbed by TGA molecularly imprinted polymer and then purified by short-path molecular distillation. S7. Membrane distillation and electrodialysis salt recovery: saline wastewater is concentrated by membrane distillation and purified by electrodialysis to recover food-grade salt and reuse pure water; S8, Digital Twin Intelligent Control System: Optimizes the entire process operation parameters in real time through a digital twin model.

2. The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to claim 1, characterized in that, The microchannel reactor in S2 has a channel size of 50-100 μm, a hydrochloric acid dropping rate of 10-15 L / h for the first acidification, a hydrochloric acid dropping rate of 8-12 L / h for the second acidification, and a reaction temperature of 20-25℃.

3. The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to claim 1, characterized in that, The composite antioxidant in S2 is composed of sodium L-cysteine, sodium bisulfite, and EDTA-2Na in a mass ratio of 2:1:0.5, and the amount added is 0.05-0.1% of the total mass of the acidification solution.

4. The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to claim 1, characterized in that, The ultrasonic pretreatment parameters in S3 are 18-22kHz, 250-350W, and 8-12min. The silica-coated iron oxide magnetic nano-flocculator has a particle size of 40-60nm and an addition amount of 50-80mg / L. The external magnetic field strength is 0.25-0.35T, and the separation time is 6-10min.

5. The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to claim 1, characterized in that, In the S4 composite extractant, the volume ratio of [BMIM][BF4] ionic liquid to MIBK is 1:4, the volume ratio of pretreatment extraction solvent to feed liquid is 1:2-1:3, the volume ratio of secondary extraction solvent to raffinate is 1:1-1:2, the dynamic ceramic membrane extractor has a membrane pore size of 200nm and a membrane surface flow rate of 0.5m / s.

6. The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to claim 1, characterized in that, The ultrasonic parameters of the ultrasonic coupled distillation column in S5 are 25-31 kHz and 450-550 W; the pre-extraction distillation vacuum degree is -0.075 to -0.085 MPa and the temperature is 63-72℃; the back-extraction distillation vacuum degree is -0.090 to -0.100 MPa and the temperature is 53-62℃.

7. The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to claim 1, characterized in that, The TGA molecularly imprinted polymer in S6 uses mercaptoacetic acid as a template molecule, ethylene glycol dimethacrylate as a crosslinking agent, adsorption temperature of 25-30℃, space velocity of 1-2 hL / (L・h), short-path molecular distillation membrane scraping speed of 1400-1600 r / min, vacuum degree of -0.09 to -0.10 MPa, and heating temperature of 60-65℃.

8. The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to claim 1, characterized in that, The S7 membrane distillation unit uses a PVDF hollow fiber membrane with a pore size of 0.15-0.25 μm, an operating temperature of 65-75℃, and a vacuum degree of -0.085 to -0.095 MPa; the electrodialysis unit has an ion exchange capacity of 1.8-2.2 mmol / g.

9. The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to claim 1, characterized in that, The digital twin model in S8 is built based on Python, integrating reaction kinetics and mass and heat transfer equations. Data is collected in real time through a TGA concentration sensor, a -SH group content sensor, and an impurity ion sensor, with a data acquisition interval of 5 seconds. The calculation formula for the digital twin model is as follows: ； In the formula: The total material change rate of TGA within the system, in mol / s; The total mass of materials within the process system, in kg; The mole fraction of TGA; Time, in seconds; The effective volume of the acidification reaction unit is expressed in cubic meters (m³). 3 ; The rate constant for the acidification reaction is... Unit m 3 / (mol·s); This represents the molar concentration of sodium thioglycolate, in mol / m³. 3 ; This refers to the molar concentration of hydrochloric acid, in mol / m³. 3 ; The mass transfer coefficient of the extraction / membrane distillation unit is expressed in m / s. The effective mass transfer area is expressed in meters (m²). 2 ; △cTGA is the TGA concentration difference during the mass transfer process, in mol / m³. 3 ; The thermal decomposition coefficient of TGA; R represents the total heat transfer of the distillation unit, in W; R is the gas constant, taken as 8.314 J / (mol·K). The operating temperature of the process unit is expressed in Kelvin (K). This refers to the loss of TGA mole fraction due to thermal decomposition. To adjust the correction coefficient, the digital twin model output is 0.8-1.2; The increment of operating parameters resulting from intelligent regulation, in mol / s.

10. The method for preparing thioglycolic acid from sodium thioglycolate and hydrogen chloride via two-stage acidification and double extraction according to claim 9, characterized in that, The S8 digital twin intelligent control system includes a physical entity layer, a sensor network layer, an edge computing and data preprocessing layer, a digital twin model core layer, an optimization decision and control layer, an actuator and command instruction layer, a data storage and historical database, and system functions and performance indicators. The physical entity layer serves as the physical carrier for the operation of the entire process unit and provides controllable objects. The sensor network layer, based on TGA concentration, -SH group content, and impurity ion sensors, collects core parameters at 5-second intervals and transmits them to the edge computing and data preprocessing layer. The edge computing and data preprocessing layer cleans, completes, and standardizes the raw data before synchronously transmitting it to the digital twin model core layer and the data storage and historical database. The digital twin model core layer, built using Python, integrates reaction kinetics and mass and heat transfer equations. It calls the preprocessed data and substitutes it into the core calculation formula to complete numerical solutions and process deviation analysis. The results are transmitted to the optimization decision and control layer. Based on the deviation, the optimization decision and control layer outputs the incremental operation parameter ΔUctrl and adjusts the control correction coefficient γ, with a value of 0.8-1.

2. This is then converted into control commands and transmitted to the actuator and command layer. The actuator and command layer drive the actuator to adjust the process parameters of the physical entity layer. At the same time, the execution results are fed back to trigger the sensor network layer to re-collect data to form a closed loop. The data storage and historical database store the entire process data and provide data support and traceability assurance. The system functions and performance indicators limit the core function boundaries and data acquisition interval ≤5 seconds and model calculation deviation ≤±3% performance thresholds.

Images