Production process of isobornyloxyethyl methacrylate

Abstract

The invention relates to the technical field of isobornyl oxyethyl methacrylate, and particularly discloses a production process of isobornyl oxyethyl methacrylate, which comprises the following steps: reacting isobornyl alcohol with ethylene oxide to obtain a target product isobornyl oxyethanol fraction; isobornyl oxyethanol and methacryloyl chloride are controlled to be subjected to an esterification reaction, and isobornyl oxyethyl acrylate is obtained; comprising the following steps: acquiring control data in the esterification reaction process of isobornyl oxyethanol and methacryloyl chloride at the current moment in real time; predicting and calculating reactant data of the current esterification reaction at the moment after the reaction is completed to obtain a regulation and control signal or a holding signal; calculating to obtain predicted regulation and control data, and judging to obtain a re-regulation and control signal or a regulation and control data qualified signal; the control of the control data at the current moment is started, the problems of side reaction increase, large concentration fluctuation and unstable conversion rate after the reaction is completed are avoided, and the conversion rate in the production process and the stability of the product purity are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of isobornyloxyethyl methacrylate, and particularly to the production process of isobornyloxyethyl methacrylate. Background Art

[0002] One of the main components of photocurable coatings is the reactive diluent. Since it has polymerizable reactive functional groups, it can participate in the photocuring crosslinking reaction, dissolve, dilute, and adjust the viscosity of the photocurable resin, reduce the VOC emissions of the photocurable coatings, and endow the photocurable coatings with excellent environmental protection characteristics. According to the different functional groups it contains, the reactive diluent can be divided into (meth)acrylate, vinyl, vinyl ether, epoxy, etc. Common acrylates have disadvantages such as strong odor and low film hardness. Isobornyl methacrylate and isobornyl acrylate, as reactive diluents and polymerization monomers, have characteristics such as high hardness, high glass transition temperature, and good optical properties, but have the problem of high brittleness. Compared with isobornyl methacrylate, the flexibility of isobornyloxyethyl methacrylate is significantly increased due to the introduction of an ethoxy fragment in its molecule, overcoming the weakness of high brittleness of isobornyl methacrylate, and at the same time having a lower viscosity and strong dilution ability.

[0003] Regarding the preparation of isobornyloxyethyl methacrylate, isobornyloxyethanol is first prepared by reacting isoborneol with ethylene oxide, and then esterification reaction is carried out with methacryloyl chloride. During the esterification reaction, since by-products will be generated during the reaction process, and heat is generated during the reaction process, in addition, the dropping rate of the reactants will affect the target product and the reaction conversion rate, etc.; In the existing reaction control, the conventional PID control or feedback regulation control method has a certain degree of hysteresis, and it is easy to have problems such as an increase in side reactions, large fluctuations in concentration, and unstable conversion rate during the production process, and it is difficult to accurately and stably control the reaction stability. Summary of the Invention

[0004] The purpose of the present invention is to provide a production process of isobornyloxyethyl methacrylate to solve the problems in the above background.

[0005] The purpose of the present invention can be achieved through the following technical solutions: The production process of isobornyloxyethyl methacrylate includes: Step 1: Isoborneol and ethylene oxide are subjected to a ring-opening addition reaction under the action of an acidic catalyst, and then filtered and distilled to obtain the target product isobornyloxyethanol fraction; Step 2: Control the esterification reaction of isobornyloxyethanol and methacryloyl chloride, that is, isobornyloxyethyl methacrylate; Specifically, the control of the esterification reaction includes: S1: Obtain the control data in real time during the esterification reaction of isobornyloxyethanol and methacryloyl chloride at the current moment st. S2: Based on the control data at the current moment st, predict and calculate the reactant data at the moment mt after the completion of the current esterification reaction, and compare it with the target qualified data to obtain a regulation signal or a hold signal. S3: Based on the regulation signal, calculate the predicted regulation data, then calculate the error value to obtain the error value data, and judge to obtain a re-regulation signal or a regulation data qualified signal. S4: Based on the regulation data qualified signal, initiate the regulation of the control data during the esterification reaction of isobornyloxyethanol and methacryloyl chloride at the current moment st.

[0006] As a further solution of the present invention: The control data includes the dropping rate Rd st , the stirring rate Sp st and the coolant flow rate Fs st ; The reactant data at the moment mt after the completion of the reaction includes the reaction temperature T mt , the by-product concentration Cby mt , the reaction conversion rate X mt and the product purity P mt .

[0007] As a further solution of the present invention: The process of predicting whether the reactant data at the moment mt after the completion of the current esterification reaction is qualified includes: Obtain the qualified historical reaction data of a complete cycle, perform normalization processing on each item of data to obtain the normalized data set , and then train the normalized reactant data set and the past control data set through the LSTM prediction model to obtain the esterification LSTM prediction model; Based on the esterification LSTM prediction model and the control data obtained at the current moment st, predict the reactant data at the moment mt after the completion of the reaction, and obtain the reactant prediction data, including the predicted value of the by-product concentration, the predicted value of the reaction conversion rate, and the predicted value of the product concentration.

[0008] As a further solution of the present invention: Compare the prediction data with the target qualified data respectively. The prediction data includes the predicted value of the by-product concentration, the predicted value of the reaction conversion rate, and the predicted value of the product concentration, and the target data includes the qualified value of the by-product concentration, the qualified value of the reaction conversion rate, and the qualified value of the product concentration; If there is one or more unqualified prediction data, generate a regulation signal; Otherwise, generate a hold signal.

[0009] As a further solution of the present invention: the prediction data being unqualified includes: The predicted value of the by-product concentration is greater than the qualified value of the by-product concentration, the predicted value of the reaction conversion rate is less than the qualified value of the reaction conversion rate, and the predicted value of the product concentration is less than the qualified value of the product concentration.

[0010] As a further solution of the present invention: the predicted regulation data includes the dropping rate regulation value Rd k , the stirring rate regulation value S k and the coolant flow rate regulation value Fs k .

[0011] As a further solution of the present invention: the calculation method of the predicted regulation data is: The normalization dataset obtained by normalizing each data based on the past dataset ; then training it through the XGBoost regulation model to obtain the esterification XGBoost regulation model; Adjust the unqualified reactant prediction data under the regulation signal to the reactant target data to obtain the regulation qualified prediction dataset ; then input the regulation qualified prediction dataset into the esterification XGBoost regulation model, and output to obtain the predicted regulation data.

[0012] As a further solution of the present invention: the calculation method of the error value is: calculating the error of each prediction data in the regulation dataset through the error function to obtain the error value data; the error value data includes the dropping rate error value, the stirring rate error value, and the coolant flow rate error value.

[0013] As a further solution of the present invention: comparing the error value data with the error reference value data respectively; among them, the error value reference includes the dropping rate error reference value, the stirring rate error reference value, and the coolant flow rate error reference value; If there is one or more data in the error value data greater than the error reference value data, generate a re-regulation signal; Otherwise, generate a regulation data qualified signal.

[0014] As a further solution of the present invention: based on the regulation data qualified signal, adjust the dropping rate Rd st , the stirring rate Sp st and the coolant flow rate Fs st at the current moment st to the dropping rate regulation value Rd k , the stirring rate regulation value S k and the coolant flow rate regulation value Fs k respectively.

[0015] The beneficial effects of the present invention: Obtained by real-time monitoring of the current control data in the preparation of isobornyloxyethyl methacrylate, and then predicting and calculating the reactant data after the control data in the current reaction process is completed, anticipating in advance whether the by-products, products and conversion rate after the reaction meet the standards. When the prediction result is unqualified, timely and rapid adjustment of the current control data is carried out in advance to avoid problems such as an increase in side reactions, large fluctuations in concentration, and unstable conversion rate after the reaction is completed, ensuring the stability of the conversion rate and the product purity in the production process, and solving the problem that the product is unqualified due to the slow abnormal regulation rate in the existing production process of isobornyloxyethyl methacrylate. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention will be further described below in conjunction with the accompanying drawings.

[0017] Figure 1 is a schematic flow chart of the esterification reaction regulation method in the present invention; Figure 2 is a schematic process diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts belong to the scope of protection of the present invention.

[0019] Embodiment 1

[0020] Please refer to Figure 1 As shown, the present invention is a production process of isobornyloxyethyl methacrylate, including the following steps: Step 1: Isoborneol and ethylene oxide are subjected to a ring-opening addition reaction under the action of an acidic catalyst, and then filtered and distilled to obtain the target product isobornyloxyethanol fraction; In the preparation process of isobornyloxyethanol, the catalyst Amberlyst-15 is used and needs to be pre-activated during addition to improve the catalytic effect during the reaction; At the same time, after this reaction, it is necessary to filter to remove the catalyst, and at the same time perform rotary evaporation to remove the solvent, and then perform vacuum distillation to recover the unreacted ethylene oxide to ensure the purity of isobornyloxyethanol; Step 2: Control the esterification reaction of isobornyloxyethanol and methacryloyl chloride. After the reaction is completed, add saturated sodium bicarbonate solution for washing, then dry with anhydrous sodium sulfate, and finally perform vacuum distillation to obtain a fraction, namely isobornyloxyethyl methacrylate; This step is an esterification reaction. Anhydrous dichloromethane is used as the solvent, and isobornyloxyethanol is added. Then, it is precooled to lower the solution temperature to 0 - 5 °C. Then, methacryloyl chloride is added while stirring. Meanwhile, triethylamine can also be added to neutralize the generated hydrochloric acid. Since the by-product in this reaction process is hydrochloric acid, and there is also unreacted methacryloyl chloride in the solution, after the reaction, the reaction solution needs to be purified, that is, to remove the by-product and unreacted methacryloyl chloride, then washed with water until neutral, then dried with anhydrous sodium sulfate, and finally rotary evaporated to remove the solvent and subjected to vacuum distillation to collect the fraction to obtain isobornyloxyethyl methacrylate.

[0021] Example 2

[0022] Based on the control of the isobornyloxyethyl methacrylate esterification reaction in the above Step 2, there is a certain degree of lag in the conventional PID control or feedback regulation control method, and problems such as an increase in side reactions, large fluctuations in concentration, and unstable conversion rate are likely to occur during the production process. Therefore, during this step of the production process, it is necessary to monitor and predict the reaction conditions to better accurately predict subsequent reaction abnormalities and perform pre-regulation to ensure the stability of the conversion rate and product purity during the production process; Refer to Figure 2 As shown, based on this, this example provides a monitoring, prediction and regulation scheme for the esterification reaction in the preparation of isobornyloxyethyl methacrylate, that is, an esterification reaction regulation method, which specifically includes the following steps: S1: Real-time obtain the control data during the esterification reaction of isobornyloxyethanol and methacryloyl chloride at the current moment st, including the dropping rate Rd st , stirring rate Sp st and coolant flow rate Fs st ; S2: Based on the control data at the current moment st, predict whether the reactant data at the moment mt after the completion of the current esterification reaction is qualified, and obtain a regulation signal or a hold signal; among them, the reactant data at the moment mt after the completion of the reaction includes the reaction temperature T mt , by-product concentration Cby mt , reaction conversion rate X mt and product purity P mt ; Among them, during the process of this esterification reaction, the current moment st and the moment mt after the completion of the reaction should be the reaction moment data and the reaction completion data corresponding to this section of the reaction moment data. Since the reaction process takes a certain amount of time, the control data at the current moment st and the reactant data at the moment mt after the completion of the reaction are relative, and the difference between mt and st should be the time required for this esterification reaction process; Specifically, obtain qualified historical reaction data for a complete cycle, and obtain the reactant data at each time node n at a unit time interval t0, including the reaction temperature T n , the by-product concentration Cby n , the reaction conversion rate X n and the product concentration P n ; Meanwhile, obtain the control data corresponding to the reactant data at each time point n, including the dropping rate Rd n , the stirring rate S n and the coolant flow rate Fs n ; where n is 1, 2, 3...; In the reactant data at time node n in the historical reaction data and the control data corresponding to the reactant data at each time point n, the data during the reaction process are matched, that is, the time difference between the two is adjusted to maintain time correspondence for better training; It should be noted that during the process of obtaining reactant data, the reaction temperature of the reactants during the esterification reaction is obtained through a temperature sensor, the by-product concentration is obtained in real-time through on-line liquid chromatography (HPLC), the reaction conversion rate is calculated by monitoring the product concentration data on-line, and the product purity is detected through gas chromatography (GC); during the process of obtaining control data, the dropping rate and the coolant flow rate are obtained through a flow meter, and the stirring rate is obtained through the stirring control system; Based on the past data within the completed time points, that is, the reactant data and control data at each completed time point, and re-label them, which are respectively recorded as the reactant data set and the past control data set, including the reaction temperature set J T , the by-product concentration set J Cby , the reaction conversion rate set J X and the product concentration set J P , the dropping rate set J Rd , the stirring rate set J S and the coolant flow rate set J Fs , and fit several groups of past data to obtain the past data set W; Among them, ; ; ; ; ; ; ; ; m is the number of training data, and its value is a positive integer. It should be noted that the larger the value of m, the more past data there is, and the more accurate the prediction will be. Based on the past dataset W, normalization processing is performed on each item of data to obtain the normalized dataset , including the reactant dataset and the past control dataset. During the process of training and predicting each item of data, normalization calculation is used to map data with different dimensions to a unified numerical range, which helps for better subsequent prediction. Then, the esterification LSTM prediction model is obtained by training the normalized reactant dataset and the past control dataset through the LSTM prediction model. Using the LSTM prediction model can accurately process non-linear data in the time series and has strong prediction ability. Then, based on the control data at the current moment st obtained by the esterification LSTM prediction model, the reactant data at the moment mt after the reaction is completed is predicted, and the reactant prediction data is obtained, including the predicted value of the by-product concentration, the predicted value of the reaction conversion rate, and the predicted value of the product concentration, to obtain the prediction dataset Y. , where mt is the moment after the reaction is completed, and the time difference between mt and st can be 10 seconds, 1 minute, or 3 minutes, etc. The specific time difference is determined based on the real-time reaction situation; Cby mt is the predicted value of the by-product concentration at the moment after the reaction is completed, X mt is the predicted value of the reaction conversion rate at the moment after the reaction is completed, P mt is the predicted value of the product concentration at the moment after the reaction is completed. It should be noted that during the initial training process of the esterification LSTM prediction model, the loss value calculated by the esterification LSTM prediction model through the loss function should be less than the loss threshold; among them, the loss function is , β is the loss value, N is the number of training samples, i is the sample time point, is the predicted value at the sample time point, is the true value at the sample time point. For example: If the number of training samples is N, after the esterification LSTM prediction model is trained, the reaction conversion rate at the time point i is predicted, and the predicted value of the reaction conversion rate obtained is YX i * , and the actual reaction conversion rate is YX i ; At this time, the loss value of the predicted value of the reaction conversion rate in the prediction dataset Y can be calculated through the loss function to obtain the reaction conversion rate loss value β. If the reaction conversion rate loss value β is less than the reaction conversion rate loss threshold, it means that the esterification LSTM prediction model can meet the requirements in the prediction of the reaction conversion rate. The prediction of other parameters such as by-product concentration and product concentration follows the same principle; Finally, the predicted data is compared with the target qualified data respectively. The predicted data includes the predicted value of by-product concentration, the predicted value of reaction conversion rate, and the predicted value of product concentration. The target data includes the qualified value of by-product concentration, the qualified value of reaction conversion rate, and the qualified value of product concentration. The target qualified data is the required value in the production process of isobornyloxyethyl methacrylate, which is set by technicians based on production requirements; If there is one or more unqualified predicted data, a regulation signal is generated; Otherwise, a holding signal is generated to maintain the reaction and continue the prediction monitoring; Among them, the situations where the predicted data is unqualified include: The predicted value of by-product concentration is greater than the qualified value of by-product concentration, the predicted value of reaction conversion rate is less than the qualified value of reaction conversion rate, and the predicted value of product concentration is less than the qualified value of product concentration; S3: Based on the regulation signal, calculate the predicted regulation data and determine the re-regulation signal or the regulation data qualified signal; Among them, the predicted regulation data includes the dropping rate regulation value Rd k 、the stirring rate regulation value S k and the coolant flow rate regulation value Fs k ; Specifically: the normalized data set obtained by normalizing each data based on the past data set ; Then, it is trained by the XGBoost regulation model to obtain the esterification XGBoost regulation model. During the training process of the esterification XGBoost regulation model, the mean square error value of the training data should be less than the error value set by technicians during the production process; Using the XGBoot regulation model for training can effectively adapt to sequence data in multiple dimensions. It adopts the decision tree ensemble method sequence, which is more suitable for data regulation between different characteristic variables; Then, the unqualified reactant predicted data in the predicted data set Y under the regulation signal is adjusted to the reactant target data. The reactant target data is the required data in the production reaction process and is the preset value of technicians to obtain the regulated qualified predicted data set ; Then, the regulated qualified predicted data set is input into the esterification XGBoost regulation model, and the predicted regulation data is output to further obtain the regulation data set Q, including the dropping rate regulation value Rd k 、the stirring rate regulation value S k and the coolant flow rate regulation value Fs k , where ; The target data of the reactants should be preset qualified values of the by-product concentration, reaction conversion rate, and product purity. The unqualified predicted reactant data are the predicted reactant data that do not meet the requirements.

[0023] Then, the error of each predicted data in the regulation dataset is calculated through the error function to obtain the error value data; the error value data includes the dropping rate error value, stirring rate error value, and coolant flow rate error value. Among them, the error calculation formula is , ω is the error value, U is the number of training samples, j is the sample time point, is the predicted value at the sample time point, is the true value at the sample time point; the principle of the error calculation formula is similar to the calculation method of the loss function, and no example is given here. And the error value data is compared with the error reference value data respectively; among them, the error reference value includes the dropping rate error reference value, stirring rate error reference value, and coolant flow rate error reference value; the error reference value is set by the technical personnel based on production requirements. If one or more of the error value data are greater than the error reference value data, a re-regulation signal is generated. Otherwise, a regulation data qualified signal is generated. S4: Based on the regulation data qualified signal, start the regulation of the control data during the esterification reaction of isobornyloxyethanol and methacryloyl chloride at the current moment st. That is, the dropping rate Rd st , stirring rate Sp st and coolant flow rate Fs st at the current moment st are respectively adjusted to the dropping rate regulation value Rd k , stirring rate regulation value S k and coolant flow rate regulation value Fs k ; It is obtained by real-time monitoring of the current control data in the preparation of isobornyloxyethyl methacrylate, and then the reactant data after the completion of the control data in the current reaction process is predicted and calculated to anticipate in advance whether the by-products, products, and conversion rates after the reaction meet the standards. When the prediction result is unqualified, the current control data is adjusted in a timely and rapid manner in advance, avoiding problems such as increased side reactions, large concentration fluctuations, and unstable conversion rates after the reaction is completed, ensuring the stability of the conversion rate and product purity during the production process, and solving the problem of unqualified products caused by slow abnormal regulation rate in the existing production process of isobornyloxyethyl methacrylate.

[0024] The above has described in detail an embodiment of the present invention, but the above content is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made in accordance with the scope of the application of the present invention shall still fall within the scope covered by the patent of the present invention.

Claims

1. The production process of isobornyloxyethyl methacrylate, characterized in that, Including: Step 1: Perform a ring-opening addition reaction between isoborneol and ethylene oxide under the action of an acidic catalyst, followed by filtration and distillation to obtain the target product, isobornyloxyethanol fraction; Step 2: Control the esterification reaction between isobornyloxyethanol and methacryloyl chloride to obtain isobornyloxyethyl methacrylate; Specifically, the control of the esterification reaction includes: S1: Obtain in real time the control data during the esterification reaction between isobornyloxyethanol and methacryloyl chloride at the current moment st; S2: Based on the control data at the current moment st, predict and calculate the reactant data at the moment mt after the completion of the current esterification reaction, and compare it with the target qualified data to obtain a regulation signal or a hold signal; S3: Based on the regulation signal, calculate the predicted regulation data, then calculate the error value to obtain the error value data, and judge to obtain a re-regulation signal or a regulation data qualified signal; S4: Based on the regulation data qualified signal, initiate the regulation of the control data during the esterification reaction between isobornyloxyethanol and methacryloyl chloride at the current moment st.

2. The production process of isobornyloxyethyl methacrylate according to claim 1, characterized in that, The control data includes the droplet acceleration rate Rd st , the stirring rate Sp st and the coolant flow rate Fs st ; Reactant data at the moment mt after the reaction is completed, including the reaction temperature T mt , by-product concentration Cby mt , reaction conversion rate X mt and product purity P mt .

3. The production process of isobornyloxyethyl methacrylate according to claim 2, characterized in that, The process of predicting whether the reactant data at the moment mt after the completion of the current esterification reaction is qualified includes: Obtain qualified historical reaction data for a complete cycle, perform normalization processing on each item of data, and obtain normalized data sets respectively , and then use the LSTM prediction model to train the normalized reactant data set and the past control data set to obtain an esterification LSTM prediction model; Predict the reactant data at the moment mt after the completion of the reaction through the esterification LSTM prediction model based on the control data obtained at the current moment st, and obtain the predicted reactant data, including the predicted value of the by-product concentration, the predicted value of the reaction conversion rate, and the predicted value of the product concentration.

4. The production process of isobornyloxyethyl methacrylate according to claim 3, characterized in that, Compare the predicted data with the target qualified data respectively. The predicted data includes the predicted value of the by-product concentration, the predicted value of the reaction conversion rate, and the predicted value of the product concentration, and the target data includes the qualified value of the by-product concentration, the qualified value of the reaction conversion rate, and the qualified value of the product concentration; If there is one or more unqualified predicted data, generate a regulation signal; Otherwise, generate a hold signal.

5. The production process of isobornyloxyethyl methacrylate according to claim 4, characterized in that, The unqualified predicted data includes: The predicted value of the by-product concentration is greater than the qualified value of the by-product concentration, the predicted value of the reaction conversion rate is less than the qualified value of the reaction conversion rate, and the predicted value of the product concentration is less than the qualified value of the product concentration.

6. The production process of isobornyloxyethyl methacrylate according to claim 5, characterized in that, The predicted regulation data includes the droplet acceleration rate regulation value Rd k , the stirring rate regulation value S k and the coolant flow rate regulation value Fs k .

7. The production process of isobornyloxyethyl methacrylate according to claim 6, characterized in that, The calculation method of the predicted regulation data is: Normalized dataset obtained by normalizing each data based on past datasets ; Then, it is trained by the XGBoost regulation model to obtain the esterification XGBoost regulation model; Adjust the unqualified reactant prediction data under the regulation signal to the target reactant data to obtain a regulation qualified prediction data set ; Then input the qualified prediction dataset for regulation into the esterification XGBoost regulation model, and the predicted regulation data is output.

8. The production process of isobornyloxyethyl methacrylate according to claim 7, characterized in that, The calculation method of the error value is: Calculate the error of each predicted data in the regulation data set through an error function to obtain the error value data; the error value data includes the dropping rate error value, the stirring rate error value, and the coolant flow rate error value.

9. The production process of isobornyloxyethyl methacrylate according to claim 8, characterized in that, Compare the error value data with the error reference value data respectively; among them, the error reference value includes the dropping rate error reference value, the stirring rate error reference value, and the coolant flow rate error reference value; If there is one or more data in the error value data greater than the error reference value data, generate a re-regulation signal; Otherwise, generate a regulation data qualified signal.

10. The production process of isobornyloxyethyl methacrylate according to claim 9, characterized in that, Based on the qualified signal of the regulation data, the dropping rate Rd st , stirring rate Sp st and coolant flow rate Fs st are respectively adjusted to the dropping rate regulation value Rd k , stirring rate regulation value S k and coolant flow rate regulation value Fs k .

Images