A method for treating waste gas in a barium chloride production process

Abstract

The present application relates to waste gas treatment technical field, specifically disclose a kind of waste gas treatment method in barium chloride production process, comprising the following steps: S1, the mixed waste gas containing hydrogen chloride and carbon dioxide is carried out dust removal and moisture removal pretreatment, then pressurization;S2, mixed waste gas is passed into the bubble column reactor filled with bifunctional ionic liquid, so that hydrogen chloride is absorbed, generate dimethyl carbonate and chloromethane;S3, gas-liquid mixture is carried out gas-liquid separation, obtain the crude product gas containing dimethyl carbonate, chloromethane, unreacted methanol and carbon dioxide, and recycling bifunctional ionic liquid absorption catalyst liquid;S4, crude product gas is sequentially condensed, rectification, separation obtains dimethyl carbonate product, recycling unreacted methanol returns step S2 and is used circularly, carbon dioxide part returns step S2 and adjusts concentration, part is discharged after purification.The method can be treated HCl and CO2, realize waste high-value conversion.

Description

Technical Field

[0001] This invention relates to the field of waste gas treatment technology, and in particular to a method for treating waste gas during barium chloride production. Background Technology

[0002] Barium chloride is an important inorganic chemical product widely used in steel heat treatment, ceramics, electronics, and chemical reagents. Its mainstream production processes (such as the hydrochloric acid-barium sulfide process and the barium carbonate-hydrochloric acid process) inevitably generate mixed waste gases containing hydrogen chloride (HCl) and carbon dioxide (CO2). Direct emission of these waste gases will severely pollute the environment; HCl will form acid rain, and CO2 is a major greenhouse gas. Therefore, effective treatment of these two waste gases is a key aspect of the green upgrading of barium chloride production.

[0003] Currently, the treatment of waste gas from barium chloride production often employs a separate treatment approach, which has several drawbacks. For hydrogen chloride waste gas, the industry commonly uses alkaline absorption (such as sodium hydroxide or lime slurry) to generate sodium chloride or calcium chloride solution. While this method can remove HCl, it produces a large amount of high-salinity wastewater, resulting in high subsequent treatment costs and failing to effectively recover chlorine resources, easily causing secondary pollution. For carbon dioxide waste gas, common treatment methods are direct high-altitude emission or simple alkaline absorption and solidification. The latter also produces carbonate waste residue with low resource value and poses a risk of carbon leakage. Existing technologies also include processes for capturing CO2 and using it to produce barium carbonate, but these processes are lengthy and not coordinated with the HCl treatment stage.

[0004] In recent years, although some studies have explored the conversion of CO2 into chemicals such as methanol and formic acid, or the catalytic oxidation of HCl into chlorine, these pathways typically require high-purity raw materials, harsh reaction conditions (such as high temperature and high pressure), or precious metal catalysts. Furthermore, they are not well-matched with the waste gas composition, concentration, and fluctuation characteristics in barium chloride production scenarios, making it difficult to meet the economic and reliability requirements for continuous industrial production. Therefore, developing an efficient, economical, and environmentally friendly integrated technology that can synergistically treat HCl and CO2, achieve high-value conversion of waste, and adapt to actual production fluctuations has become a pressing technical challenge in this field. Summary of the Invention

[0005] To address the problems of existing technologies, such as the separate treatment of hydrogen chloride and carbon dioxide waste gases, complex processes, low resource utilization, and easy generation of secondary pollution (such as saline wastewater), this invention provides a waste gas treatment method in the barium chloride production process.

[0006] The technical solution adopted in this invention is:

[0007] A method for treating waste gas during barium chloride production includes the following steps:

[0008] S1. The mixed waste gas containing hydrogen chloride and carbon dioxide from the barium chloride production system is pretreated by dust removal and dehumidification, and then the pretreated mixed waste gas is pressurized to 1.5-3.0 MPa;

[0009] S2. The pressurized mixed waste gas is introduced into a bubble column reactor filled with bifunctional ionic liquid. The reaction temperature is controlled at 80-120℃, so that the hydrogen chloride in the mixed waste gas is absorbed by the bifunctional ionic liquid. At the same time, under the catalytic action of the bifunctional ionic liquid, the absorbed hydrogen chloride reacts with carbon dioxide in the waste gas and continuously introduced methanol to generate dimethyl carbonate and chloromethane.

[0010] S3. The gas-liquid mixture from the bubble column reactor is separated into gas and liquid to obtain a crude product gas rich in dimethyl carbonate, chloromethane, unreacted methanol and excess carbon dioxide, as well as a bifunctional ionic liquid absorption catalyst for recycling.

[0011] S4. The crude product gas is sequentially condensed and distilled to separate high-purity dimethyl carbonate product. Unreacted methanol is recovered and returned to step S2 for recycling. Chloromethane and carbon dioxide are separated from the remaining tail gas through cryogenic separation. Part of the carbon dioxide is returned to step S2 to adjust the concentration, and the rest is purified before being discharged.

[0012] Preferably, in step S2, the bifunctional ionic liquid is a functionalized ionic liquid based on organic base cations and chloride or amino acid anions, with the general formula: , wherein the cation It is one of 1-butyl-3-methylimidazolium, 1-hydroxyethyl-3-methylimidazolium, and tetrabutylphosphine; the anion It is one of chloride ions, glycine ions, and proline ions.

[0013] Preferably, in step S2, the gas-liquid residence time ratio in the bubbling tower reactor is controlled to be 0.5-2:1, and the molar ratio of methanol continuously introduced into the bifunctional ionic liquid to the theoretically calculated required hydrogen chloride is 2:1 to 5:1.

[0014] Preferably, in step S2, the operating temperature of the bubble column reactor is... Operating pressure and methanol feed flow rate The settings are dynamically optimized by an online adaptive neural network control system, which performs the following steps:

[0015] S21. Real-time acquisition of the volumetric flow rate of the mixed exhaust gas entering the bubbling tower reactor. Hydrogen chloride concentration carbon dioxide concentration And the mole fraction of dimethyl carbonate in the outlet gas of the bubble column reactor. ;

[0016] S22, Collect the data , , As input, it is fed into a pre-trained deep neural network model, which outputs the optimal operating parameter settings for the current moment, including the optimal operating temperature. Optimal operating pressure Optimal methanol feed flow rate ;

[0017] S23. The optimal operating parameter settings are sent to the temperature controller, pressure regulating valve and methanol feed pump of the bubbling tower reactor to adjust the reactor to operate in the optimal state.

[0018] Preferably, the deep neural network model is a feedforward neural network with residual connections, and its overall architecture includes:

[0019] The input layer contains three neurons, each corresponding to a normalized input feature. ,in, , , , and These are the mean and standard deviation of the corresponding features on the training dataset, respectively;

[0020] The hidden layer contains 5 fully connected layers with 128, 256, 256, 128, and 64 neurons respectively. Except for the 3rd layer, each layer is followed by a batch normalization layer and a ReLU activation function. Residual connections are set between the 2nd and 4th layers and between the 1st and 5th layers.

[0021] The output layer contains three neurons, each corresponding to a normalized optimal setpoint for the operating parameters. The output value is mapped to using the Sigmoid activation function. The interval is then denormalized to obtain the actual set value. ,in, The permitted range of operations for the corresponding feature. The optimal operating parameter settings for the corresponding features output by the deep neural network model at the current time step. The optimal operating parameter setting value for the normalization of the corresponding feature.

[0022] Preferably, the deep neural network model is trained through the following process:

[0023] Different inlet conditions of the bubble column reactor and operating parameters The system operates under the following combination of conditions, and the mole fraction of dimethyl carbonate in the corresponding outlet gas is collected. To form a dataset ,in, ;

[0024] Define benefit function ,in, , and These are the weighting coefficients. and For reference flow rate and temperature; for each set of inlet conditions, the search makes The largest combination of operating parameters is taken as the optimal output label corresponding to that set of input conditions;

[0025] The loss function is to minimize the mean squared error (MSE) between the neural network's predicted parameters and the optimal label parameters. The Adam optimizer is used for training.

[0026] Preferably, in step S3, the bifunctional ionic liquid absorption catalyst obtained for recycling is first passed through a vacuum flash tank before returning to the bubble column reactor to remove trace amounts of dissolved water and methanol under conditions of 100-120°C and 5-15 kPa absolute pressure, and then passed through an adsorption tower equipped with 3A molecular sieve for deep drying.

[0027] Preferably, in step S4, the specific operation of partially returning to step S2 to adjust the concentration is as follows: the tail gas that needs to be returned to step S2 is pressurized and mixed with the fresh mixed waste gas from step S1 in a certain proportion, so that the molar fraction of carbon dioxide in the mixed waste gas entering the bubble tower reactor in step S2 is maintained at 30%-50%.

[0028] Preferably, in step S2, the gas phase distributor of the bubble column reactor is made of sintered metal or ceramic material with micron-sized pores, and its average pore size is... The thickness is 5-50 μm, and the reactor is filled with structured packing or stepped ring packing. The bifunctional ionic liquid is distributed in a film-like manner on the surface of the packing. The apparent gas velocity of the mixed waste gas is... The circulation rate of the liquid-phase ionic liquid is controlled at 0.05-0.15 m / s. The velocity should be controlled within the range of 0.005-0.02 m / s and should satisfy the following relationship. To optimize the gas-liquid mass transfer interface, among which, To optimize the coefficients, the value range is 8-15. and These represent the densities of the gas-phase mixture and the liquid-phase ionic liquid under operating conditions, respectively. and These represent the viscosity of the gas-phase mixture and the viscosity of the liquid-phase ionic liquid under operating conditions, respectively.

[0029] The beneficial effects of this invention are:

[0030] 1. Using a bifunctional ionic liquid as a medium, the absorption and fixation of HCl and the catalytic conversion of CO2 are simultaneously completed in a single reaction unit (bubbling tower reactor). The two waste gases are not simply neutralized or sealed, but are used as raw materials to directly synthesize dimethyl carbonate (DMC) and chloromethane, which have high market value, under mild conditions (80-120℃, 1.5-3.0MPa). This transforms pollutants into chemical products and solves the fundamental defects of traditional methods, such as low resource utilization and easy generation of secondary pollution.

[0031] 2. By seamlessly coupling waste gas purification with chemical production, the process flow is significantly shortened. Compared to the traditional two independent systems of "alkali washing and deacidification + CO2 capture and storage", this invention only requires one core reaction-separation device, which significantly reduces equipment investment, floor space and operating energy consumption. Moreover, the product DMC is an important green solvent and organic synthesis intermediate, and chloromethane is a key raw material for the synthesis of bulk chemicals such as organosilicon. Its sales revenue can fully cover or even exceed the cost of waste gas treatment, turning environmental protection investment into a profit-making link. At the same time, it eliminates the generation of saline wastewater and waste residue from the source, resulting in outstanding environmental benefits.

[0032] 3. In bifunctional ionic liquids, the cationic portion enhances gas capacity through physical dissolution and charge interaction, while the anionic portion serves as an HCl absorption site and a CO2 activation catalytic center. This design allows absorption and catalytic reactions to proceed synergistically at the same active site, resulting in a direct reaction pathway, high efficiency, and extremely low vapor pressure and thermal stability. Furthermore, the ionic liquid itself is recyclable and can maintain its activity for a long time through a matching vacuum flash evaporation-molecular sieve regeneration unit, resulting in low operating costs.

[0033] 4. A CO2 tail gas recirculation loop is set up to return unreacted CO2 to the reactor inlet. By adjusting the CO2 concentration in the mixed gas (maintaining it at 30%-50%), the reaction kinetic environment is effectively stabilized, the overall CO2 utilization rate is improved, and fluctuations in the upstream waste gas concentration are buffered. Furthermore, by introducing an online adaptive neural network control system, the reaction temperature, pressure, and methanol feed rate are dynamically optimized in real time based on the flow rate and concentration of the waste gas feed. Through a pre-trained deep learning model, the system automatically finds and maintains the optimal economic efficiency under the current operating conditions, giving the entire process strong anti-interference capabilities and operational robustness. Even when facing fluctuations in barium chloride production load, it can still maintain high conversion rate and low material and energy consumption. Attached Figure Description

[0034] Figure 1This is a schematic diagram of the waste gas treatment method during the barium chloride production process in an embodiment of the present invention;

[0035] Figure 2 This is a schematic diagram illustrating the dynamic optimization process of the online adaptive neural network control system in an embodiment of the present invention. Detailed Implementation

[0036] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0037] Example

[0038] A method for treating waste gas during barium chloride production, such as Figure 1 As shown, it includes the following steps:

[0039] S1. Waste gas pretreatment and compression: The mixed waste gas containing hydrogen chloride and carbon dioxide from the barium chloride production system is pretreated by dust removal and dehumidification, and then the pretreated mixed waste gas is pressurized to 1.5-3.0 MPa;

[0040] S2. Synergistic absorption and catalytic reaction of ionic liquid: The pressurized mixed waste gas is passed into a bubble column reactor filled with bifunctional ionic liquid. The reaction temperature is controlled at 80-120℃, so that the hydrogen chloride in the mixed waste gas is absorbed by the bifunctional ionic liquid. At the same time, under the catalytic action of the bifunctional ionic liquid, the absorbed hydrogen chloride reacts with carbon dioxide in the waste gas and continuously introduced methanol to generate dimethyl carbonate and chloromethane.

[0041] S3. Separation of reaction products: The gas-liquid mixture from the bubble column reactor is separated into gas and liquid to obtain a crude product gas rich in dimethyl carbonate, chloromethane, unreacted methanol and excess carbon dioxide, as well as a bifunctional ionic liquid absorption catalyst for recycling.

[0042] S4. Product Refining: The crude product gas is sequentially condensed and distilled to separate high-purity dimethyl carbonate product. Unreacted methanol is recovered and returned to step S2 for recycling. Chloromethane and carbon dioxide are separated from the remaining tail gas through cryogenic separation. Part of the carbon dioxide is returned to step S2 to adjust the concentration, and the rest is purified before being discharged.

[0043] By employing an integrated ionic liquid with dual absorption and catalytic functions, the absorption and fixation of hydrogen chloride and the catalytic conversion of carbon dioxide are simultaneously completed within a single reaction unit. Pressurization facilitates the dissolution of gases in the ionic liquid and shifts the reaction equilibrium towards the products. The ionic liquid with its specific structure can reversibly absorb HCl through acid-base interactions and also activate CO2 molecules. At the reaction temperature, the absorbed HCl, dissolved CO2, and methanol undergo an esterification-carbonylation coupling reaction. This directly converts two types of waste gas pollutants into high-value dimethyl carbonate, achieving waste-to-resource transformation. The highly integrated process significantly reduces equipment investment and floor space requirements, and avoids the environmental problems associated with large amounts of difficult-to-treat saline wastewater generated by traditional alkaline absorption methods, thereby improving resource utilization and environmental benefits from the source.

[0044] To achieve both efficient HCl absorption and CO2 conversion, thereby improving reaction efficiency, in step S2, the bifunctional ionic liquid is a functionalized ionic liquid based on organic base cations and chloride or amino acid anions, with the general formula: , wherein the cation It is one of 1-butyl-3-methylimidazolium, 1-hydroxyethyl-3-methylimidazolium, and tetrabutylphosphine; the anion It is one of chloride ions, glycine ions, and proline ions.

[0045] Cations such as 1-butyl-3-methylimidazolium, 1-hydroxyethyl-3-methylimidazolium, and tetrabutylphosphonium exhibit good chemical stability and gas solubility regulation capabilities. Anions such as chloride ions can directly participate in the reaction, while glycine and proline ions contain both basic sites (amino groups) that can bind to H⁺ and active sites (carboxyl groups) that can activate CO₂. Therefore, bifunctional ionic liquids can not only physically dissolve and chemically absorb HCl, but their anions can also act as nucleophiles or basic catalysts, promoting the reaction of CO₂ with methanol and HCl derivatives (such as in-situ generated CH₃Cl) to form DMC.

[0046] In order to optimize the hydrodynamic conditions and reactant ratio in the reactor to balance gas absorption, reaction kinetics and mass transfer efficiency, and to prevent low conversion rate due to insufficient residence time or waste of raw materials and increase of by-products due to improper ratio, in step S2, the gas-liquid residence time ratio in the bubble column reactor is controlled to be 0.5-2:1, and the molar ratio of methanol continuously introduced into the bifunctional ionic liquid to the theoretically calculated required hydrogen chloride is 2:1 to 5:1.

[0047] The gas-liquid residence time ratio directly affects the dissolution and reaction time of the reactant gases (HCl, CO2) in the liquid phase (ionic liquid). An appropriate ratio ensures that the gases have sufficient time to be absorbed and reacted, while maintaining the reactor's throughput. Excess methanol can shift the reaction equilibrium to the right, increasing the conversion rates of HCl and CO2; however, excessive excess will increase the energy consumption of subsequent separation. The defined molar ratio range (2:1 to 5:1) is an optimized range derived from the stoichiometry of this coupled reaction and engineering experience, achieving the best balance between reaction efficiency and operational economy while ensuring high conversion rates.

[0048] To avoid fluctuations in the upstream barium chloride production process leading to unstable waste gas concentration and flow rate, which could result in decreased processing efficiency, product quality fluctuations, or increased energy consumption in the fixed-parameter reaction system, the operating temperature of the bubbling tower reactor is adjusted in step S2. Operating pressure and methanol feed flow rate Dynamically optimized settings are achieved by an online adaptive neural network control system, such as... Figure 2 As shown, the control system performs the following steps:

[0049] S21. Data Acquisition: Real-time acquisition of the volumetric flow rate of the mixed waste gas entering the bubbling tower reactor. Hydrogen chloride concentration carbon dioxide concentration And the mole fraction of dimethyl carbonate in the outlet gas of the bubble column reactor. ;

[0050] S22. Optimize calculations: Calculate the collected data... , , As input, it is fed into a pre-trained deep neural network model, which outputs the optimal operating parameter settings for the current moment, including the optimal operating temperature. Optimal operating pressure Optimal methanol feed flow rate ;

[0051] S23. Execution control: The optimal operating parameter setpoints are sent to the temperature controller, pressure regulating valve and methanol feed pump of the bubbling tower reactor to adjust the reactor to operate in the optimal state.

[0052] By sensing changes in exhaust gas inlet conditions in real time, the system uses a neural network model to quickly predict the optimal reaction conditions (temperature, pressure, alcohol-acid ratio) and adjusts them in real time. Utilizing the powerful nonlinear fitting capability of the neural network, it learns the hidden high-dimensional mapping relationship between inlet conditions and optimal operating parameters, enabling the treatment system to have adaptive capabilities. It can automatically maintain the optimal operating point when faced with fluctuating feed, thereby maximizing the yield and purity of dimethyl carbonate in long-term operation while minimizing energy and methanol consumption, thus improving the robustness and economy of the entire method.

[0053] Specifically, the deep neural network model is a feedforward neural network with residual connections, and its overall architecture includes:

[0054] The input layer contains three neurons, each corresponding to a normalized input feature. ,in, , , , and These are the mean and standard deviation of the corresponding features on the training dataset, respectively;

[0055] The hidden layer contains 5 fully connected layers with 128, 256, 256, 128, and 64 neurons respectively. Except for the 3rd layer, each layer is followed by a batch normalization layer and a ReLU activation function. Residual connections are set between the 2nd and 4th layers and between the 1st and 5th layers.

[0056] The output layer contains three neurons, each corresponding to a normalized optimal setpoint for the operating parameters. The output value is mapped to using the Sigmoid activation function. The interval is then denormalized to obtain the actual set value. ,in, The permitted range of operations for the corresponding feature. The optimal operating parameter settings for the corresponding features output by the deep neural network model at the current time step. The optimal operating parameter setting value for the normalization of the corresponding feature.

[0057] This network employs a deep feedforward structure to learn complex patterns. Residual connections help mitigate gradient vanishing, making the deep network easier to train. Batch normalization accelerates training and improves generalization, while input / output normalization / de-normalization ensures the network operates within a stable numerical range. Through training on a large amount of historical or simulated data, the network learns the mapping from "exhaust gas state" to "optimal operation." The deep structure ensures the model has sufficient ability to fit complex process relationships, and techniques such as residual connections improve the model's training stability and prediction accuracy, thus providing reliable core algorithmic support for online optimization.

[0058] To ensure that the "optimal" operation learned by the model aligns with the economic goals of the production process, rather than simply maximizing the conversion rate, the deep neural network model is trained through the following process:

[0059] Different inlet conditions of the bubble column reactor and operating parameters The system operates under the following combination of conditions, and the mole fraction of dimethyl carbonate in the corresponding outlet gas is collected. To form a dataset ,in, ;

[0060] Define benefit function ,in, , and These are the weighting coefficients. and For reference flow rate and temperature; for each set of inlet conditions, the search makes The largest combination of operating parameters is taken as the optimal output label corresponding to that set of input conditions;

[0061] The loss function is to minimize the mean squared error (MSE) between the neural network's predicted parameters and the optimal label parameters. The Adam optimizer was used for training, with an initial learning rate of 0.001 and a cosine annealing strategy for adjustment. The training epochs were 500.

[0062] The core of the above training process is to construct an optimization problem with the goal of comprehensive economic benefit (Profit). The Profit function considers not only the yield of the main product (…). It also takes into account the consumption of major raw materials ( ) and energy consumption (usually related to temperature) Positive correlation By using offline data or simulation, the operating parameters that maximize the Profit value are found for each set of exhaust gas inlet conditions. These parameters are then used as "teacher signals" to train the network. The operating point recommended by the trained neural network model is the economically optimal point for the system under given inlet conditions, rather than simply the point with the highest chemical conversion rate. This makes the entire system closer to actual production needs and achieves a unity of technical and economic indicators.

[0063] The water and excess methanol generated during the reaction accumulate in the ionic liquid. These accumulations reduce the absorption capacity and catalytic activity of the ionic liquid and may promote side reactions. Therefore, in step S3, the bifunctional ionic liquid absorption catalyst obtained for recycling is first passed through a vacuum flash tank before returning to the bubble column reactor. Under the conditions of temperature 100-120℃ and absolute pressure 5-15kPa, the trace amounts of dissolved water and methanol are removed. Then, it is deeply dried in an adsorption tower equipped with 3A molecular sieves.

[0064] Vacuum flash evaporation utilizes the principle of lowering the boiling point of a liquid by reducing pressure, effectively removing most volatile water and methanol at relatively low temperatures with relatively low energy consumption. Subsequent 3A molecular sieve adsorption is a highly efficient deep dehydration technology; its pore size only allows water molecules to enter, further reducing the water content in the ionic liquid to extremely low levels (e.g., <50 ppm). This two-stage purification works synergistically to continuously and efficiently maintain the "dry" and "clean" state of the circulating ionic liquid, ensuring long-term stability of its absorption and catalytic performance. This avoids frequent replacements or regenerations due to ionic liquid deactivation, reduces operating costs, and guarantees long-term stable operation of the entire system.

[0065] To address the issue of decreased coupling reaction rate and reduced DMC yield due to fluctuations in CO2 concentration in the exhaust gas (especially when the concentration is low), and to ensure that the composition of the feed gas entering the reactor remains stable within the optimal reaction range, the specific operation of returning a portion of the exhaust gas to step S2 for concentration adjustment in step S4 is as follows: the exhaust gas to be returned to step S2 is pressurized and then mixed with the fresh mixed exhaust gas from step S1 in a proportional manner, so that the molar fraction of carbon dioxide in the mixed exhaust gas entering the bubble column reactor in step S2 is maintained at 30%-50%.

[0066] Since the rate and equilibrium of the coupled reaction are closely related to the partial pressure (concentration) of CO2, recycling the unreacted CO2 tail gas and mixing it with the original waste gas allows for active control of the CO2 concentration in the mixed gas entering the reactor, stabilizing it within the optimized range of 30%-50%. This ensures that the reaction always proceeds under favorable thermodynamic and kinetic conditions, improving the overall utilization rate of CO2. Furthermore, it stabilizes the reaction environment, reduces disturbances in the main reactor caused by fluctuations in feed concentration, and is beneficial for producing products with stable quality and extending catalyst life. In addition, the operation is simple and easy to implement, requiring only the addition of circulation pipelines and proportional control valves, resulting in low cost but significant effects.

[0067] In a gas-liquid-solid multiphase reaction system (where the catalyst is a functionalized ionic liquid, which can be considered quasi-homogeneous, but the reaction occurs at the gas-liquid interface), the overall reaction rate is controlled by mass transfer due to the slow mass transfer rates of HCl and CO2 in viscous ionic liquids and the limited reaction interface. Therefore, it is necessary to increase the effective gas-liquid mass transfer specific surface area and the interface renewal rate. In step S2, the gas phase distributor of the bubble column reactor is made of sintered metal or ceramic material with micron-sized pores, and its average pore size... The thickness is 5-50 μm, and the reactor is filled with inert, high specific surface area structured packing or stepped ring packing, with the bifunctional ionic liquid distributed in a film form on the surface of the packing; the apparent gas velocity of the mixed waste gas is 5-50 μm. The circulation rate of the liquid-phase ionic liquid is controlled at 0.05-0.15 m / s. The velocity should be controlled within the range of 0.005-0.02 m / s and should satisfy the following relationship. To optimize the gas-liquid mass transfer interface, among which, To optimize the coefficients, the value range is 8-15. and These represent the densities of the gas-phase mixture and the liquid-phase ionic liquid under operating conditions, respectively. and These represent the viscosity of the gas-phase mixture and the viscosity of the liquid-phase ionic liquid under operating conditions, respectively.

[0068] Regarding the best The value can be determined using the following methods:

[0069] 1. Cold model testing method:

[0070] In a cold model device with geometry similar to an actual reactor, an air-water or air-simulated ionic liquid (similar viscosity, density) system is used.

[0071] System change and Measure gas content Pressure drop and liquid phase mass transfer coefficient .

[0072] draw Follow Find the changing curve The region where a plateau or peak occurs, the region corresponding to Value, combined with the characteristic parameters at that time Substituting into the formula and working backwards, we can obtain the optimal system for this condition. value.

[0073] 2. Thermal model experiment method:

[0074] In a small-scale continuous experimental setup, real reactants (simulated waste gas containing HCl and CO2, methanol, and bifunctional ionic liquid) are used.

[0075] With other conditions fixed (temperature, pressure, concentration), the system changes and However, maintain its ratio According to different presets Values ​​(such as 8, 10, 12, 14) are used for calculation or setting.

[0076] Measurement of different Key performance indicators under the specified value: HCl conversion rate ( ), CO2 single-pass conversion rate ( DMC Selectivity (and unit energy consumption).

[0077] Using a comprehensive benefit function (such as) Using as the evaluation criterion, determine the criterion that maximizes the function. The value is the optimal value for that specific reaction system. value.

[0078] The embodiments described above are merely illustrative of specific implementations of the present invention, and while the descriptions are detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A method for treating waste gas during barium chloride production, characterized in that, Includes the following steps: S1. The mixed waste gas containing hydrogen chloride and carbon dioxide from the barium chloride production system is pretreated by dust removal and dehumidification, and then the pretreated mixed waste gas is pressurized to 1.5-3.0 MPa; S2. The pressurized mixed waste gas is introduced into a bubble column reactor filled with bifunctional ionic liquid. The reaction temperature is controlled at 80-120℃, so that the hydrogen chloride in the mixed waste gas is absorbed by the bifunctional ionic liquid. At the same time, under the catalytic action of the bifunctional ionic liquid, the absorbed hydrogen chloride reacts with carbon dioxide in the waste gas and continuously introduced methanol to generate dimethyl carbonate and chloromethane. S3. Separate the gas-liquid mixture from the bubble column reactor to obtain a crude product gas containing dimethyl carbonate, chloromethane, unreacted methanol and carbon dioxide, as well as a bifunctional ionic liquid absorption catalyst for recycling. S4. The crude product gas is sequentially condensed and distilled to separate dimethyl carbonate product. Unreacted methanol is recovered and returned to step S2 for recycling. Chloromethane and carbon dioxide are separated from the remaining tail gas by cryogenic separation. Part of the carbon dioxide is returned to step S2 to adjust the concentration, and part is discharged after purification.

2. The waste gas treatment method in the barium chloride production process according to claim 1, characterized in that, In step S2, the bifunctional ionic liquid is a functionalized ionic liquid based on organic base cations and chloride or amino acid anions, with the general formula: , wherein the cation It is one of 1-butyl-3-methylimidazolium, 1-hydroxyethyl-3-methylimidazolium, and tetrabutylphosphine; the anion It is one of chloride ions, glycine ions, and proline ions.

3. The waste gas treatment method in the barium chloride production process according to claim 1, characterized in that, In step S2, the gas-liquid residence time ratio in the bubbling tower reactor is controlled to be 0.5-2:1, and the molar ratio of methanol continuously introduced into the bifunctional ionic liquid to the theoretically calculated required hydrogen chloride is 2:1 to 5:

1.

4. The waste gas treatment method in the barium chloride production process according to claim 1, characterized in that, In step S2, the operating temperature of the bubble column reactor... Operating pressure and methanol feed flow rate The settings are dynamically optimized by an online adaptive neural network control system, which performs the following steps: S21. Real-time acquisition of the volumetric flow rate of the mixed exhaust gas entering the bubbling tower reactor. Hydrogen chloride concentration carbon dioxide concentration And the mole fraction of dimethyl carbonate in the outlet gas of the bubble column reactor. ; S22, Collect the data , , As input, it is fed into a pre-trained deep neural network model, which outputs the optimal operating parameter settings for the current moment, including the optimal operating temperature. Optimal operating pressure Optimal methanol feed flow rate ; S23. The optimal operating parameter settings are sent to the temperature controller, pressure regulating valve and methanol feed pump of the bubbling tower reactor to adjust the reactor to operate in the optimal state.

5. The waste gas treatment method in the barium chloride production process according to claim 4, characterized in that, The deep neural network model is a feedforward neural network with residual connections, and its overall architecture includes: The input layer contains three neurons, each corresponding to a normalized input feature. ,in, , , , and These are the mean and standard deviation of the corresponding features on the training dataset, respectively; The hidden layer contains 5 fully connected layers with 128, 256, 256, 128, and 64 neurons respectively. Except for the 3rd layer, each layer is followed by a batch normalization layer and a ReLU activation function. Residual connections are set between the 2nd and 4th layers and between the 1st and 5th layers. The output layer contains three neurons, each corresponding to a normalized optimal setpoint for the operating parameters. The output value is mapped to using the Sigmoid activation function. The interval is then denormalized to obtain the actual set value. ,in, The permitted range of operations for the corresponding feature. The optimal operating parameter settings for the corresponding features output by the deep neural network model at the current time step. The optimal operating parameter setting value for the normalization of the corresponding feature.

6. The waste gas treatment method in the barium chloride production process according to claim 5, characterized in that, The deep neural network model is trained through the following process: Different inlet conditions of the bubble column reactor and operating parameters The system operates under the following combination of conditions, collecting the mole fraction of dimethyl carbonate in the corresponding outlet gas. To form a dataset ,in, ; Define benefit function ,in, , and These are the weighting coefficients. and For reference flow rate and temperature; for each set of inlet conditions, the search makes The largest combination of operating parameters is taken as the optimal output label corresponding to that set of input conditions; The loss function is to minimize the mean squared error (MSE) between the neural network's predicted parameters and the optimal label parameters. The Adam optimizer is used for training.

7. The waste gas treatment method in the barium chloride production process according to claim 1, characterized in that, In step S3, the bifunctional ionic liquid absorption catalyst obtained for recycling is first passed through a vacuum flash tank before returning to the bubble column reactor. Under the conditions of temperature 100-120℃ and absolute pressure 5-15kPa, trace amounts of dissolved water and methanol are removed. Then, it is further dried in an adsorption tower equipped with 3A molecular sieves.

8. The waste gas treatment method in the barium chloride production process according to claim 1, characterized in that, In step S4, the specific operation of returning part of the exhaust gas to step S2 to adjust the concentration is as follows: the exhaust gas that needs to be returned to step S2 is pressurized and mixed with the fresh mixed exhaust gas from step S1 in a certain proportion, so that the molar fraction of carbon dioxide in the mixed exhaust gas entering the bubble column reactor of step S2 is maintained at 30%-50%.

9. The waste gas treatment method in the barium chloride production process according to claim 1, characterized in that, In step S2, the gas phase distributor of the bubble column reactor is made of sintered metal or ceramic material with micron-sized pores, and its average pore size is... The thickness is 5-50 μm, and the reactor is filled with structured packing or stepped ring packing. The bifunctional ionic liquid is distributed in a film-like manner on the surface of the packing. The apparent gas velocity of the mixed waste gas is... The circulation rate of the liquid-phase ionic liquid is controlled at 0.05-0.15 m / s. The velocity should be controlled within the range of 0.005-0.02 m / s and should satisfy the following relationship. To optimize the gas-liquid mass transfer interface, among which, To optimize the coefficients, the value range is 8-15. and These represent the densities of the gas-phase mixture and the liquid-phase ionic liquid under operating conditions, respectively. and These represent the viscosity of the gas-phase mixture and the viscosity of the liquid-phase ionic liquid under operating conditions, respectively.

Images