Method for optimizing synthesis of aryl porphyrin compound

Abstract

The invention discloses a method for optimizing synthesis of an aryl porphyrin compound. Collecting initial data; inputting the initial data into the proxy model, and training by using a Bayesian optimization algorithm to obtain an optimal reaction condition combination; and carrying out synthesis reaction according to the optimal reaction condition combination and the specific types of the aryl porphyrin compounds to obtain the aryl porphyrin compounds. Taking tetraphenylporphyrin as an example, the optimal conditions obtained by a verification experiment are as follows: the molar ratio of pyrrole to benzaldehyde is 1: 1, toluene is used as a solvent, propionic acid is used as an acidic solvent, the volume ratio of pyrrole to benzaldehyde is 1: 3, the reaction concentration is 1 mol / L, and p-toluenesulfonic acid is used as an additive, and the amount of p-toluenesulfonic acid is 5 mol%. According to the method, scientific decision support is provided under the condition of high uncertainty, the synthesis efficiency of tetraphenylporphyrin is improved, the number of experimental tests is reduced to the maximum extent, consumption of chemical reagents is reduced, and a firmer foundation is laid for research and application in related fields.

Description

Technical Field

[0001] The present invention relates to a method for synthesizing aryl porphyrin compounds, and more particularly to a method for optimizing the synthesis of aryl porphyrin compounds. Background Art

[0002] Aryl porphyrin compounds are a diverse class of macrocyclic molecules with unique electronic and optical properties that make them highly valuable in applications such as catalysis, energy storage, photodynamic therapy, and chemical sensing. These compounds consist of a porphyrin ring with aromatic substituents and can exhibit specific properties, such as enhanced light absorption, improved stability, and tunable reactivity, by varying the substituents.

[0003] Tetraphenylporphyrin (TPP) belongs to aryl porphyrin compounds. It is a large heterocyclic compound formed by the α-carbon atoms of four pyrrole subunits interconnected by methylene bridges (=CH-), and is an important compound. Currently, it is widely used in many fields such as catalysis, optoelectronic devices, medicine, and bioimaging. As a precursor for a class of metal complexes, tetraphenylporphyrin and its derivatives have attracted much attention due to their excellent optical and electronic properties. In particular, in the applications of photocatalysis and photosensitizers, the synthesis process of tetraphenylporphyrin has become a research hotspot.

[0004] Currently, the main method for synthesizing porphyrin is the Adler method, and the reaction mechanism is as follows:

[0005]

[0006] In this reaction, first, the aldehyde group is protonated, making the oxygen more electrophilic and forming an active carbocation intermediate. The protonated aldehyde group is then subjected to a nucleophilic attack by the amine on the pyrrole ring. This results in the formation of an intermediate imine ion. Since the imine ion is unstable, proton dissociation occurs and it undergoes a dehydration step to form an aziridinium ion. This three-membered ring intermediate is highly electrophilic. The aziridinium ion is unstable and is subjected to a nucleophilic attack by another amine group, and a Michael addition reaction occurs, thus forming a cyclic product. After multiple nucleophilic additions, a polycyclic amine compound is formed. Extended porphyrin compounds can be obtained by connecting different substituents to the aromatic ring. This reaction mechanism is mainly based on acylation and cyclization reactions, involving multiple steps, and the reaction conditions of each step will affect the final yield and purity.

[0007] In recent years, with the development of artificial intelligence (AI) and machine learning technologies, these emerging technologies have provided new ideas and methods for chemical synthesis. Currently, the synthesis process of tetraphenylporphyrin faces challenges in environmental protection and sustainable development. Traditional chemical synthesis methods often require a large number of experiments to determine the final synthesis conditions, during which there will be losses of chemical reagents, posing potential risks to the environment. In recent years, the proposal of the concept of green chemistry has prompted researchers to explore more environmentally friendly and sustainable synthesis routes. Using artificial intelligence technology to optimize reaction conditions can effectively improve reaction efficiency, reduce unnecessary reagent consumption, and reduce the generation of harmful by-products, meeting the requirements of green chemistry. Although some current studies have achieved preliminary results using artificial intelligence technology in organic synthesis, the specialized optimization research on tetraphenylporphyrin has not been perfected. With the increasing demand for tetraphenylporphyrin and its derivatives, it has become particularly important to develop efficient, economical, and environmentally friendly synthesis methods. Summary of the Invention

[0008] In order to improve the yield and purity of arylporphyrin compounds, the present invention provides a method for optimizing the synthesis of arylporphyrin compounds. By using optimization algorithms such as Bayesian combined with surrogate models, the reaction parameters are dynamically adjusted to continuously optimize the synthesis process, and finally the optimal method for synthesizing arylporphyrin compounds is obtained. Compared with Adler's method, while increasing the yield, it minimizes the generation of by-products, further improves the purity of the final product, and improves the repeatability of experiments, thereby enhancing the possibility of industrial production of arylporphyrin compounds.

[0009] To solve the problems of the existing technology, the technical solution adopted by the present invention is as follows:

[0010] A method for optimizing the synthesis of arylporphyrin compounds, comprising the following steps:

[0011] Step 1, collect initial data, wherein the initial data includes the molar ratio between reactants, additive concentration, types and amounts of solvents used for reactants, types and amounts of acidic solvents, reaction temperature, reaction time, distillation;

[0012] Step 2, input the initial data into the surrogate model and use the Bayesian optimization algorithm for training to obtain the optimal combination of reaction conditions;

[0013] Step 3, according to the optimal combination of reaction conditions in Step 2, conduct a synthesis reaction according to the specific type of arylporphyrin compound to obtain the arylporphyrin compound.

[0014] Preferably, the source of the reaction parameters in Step 1 is data obtained from literature research and conventional experiments.

[0015] Preferably, the specific steps of training in step 2 are as follows: Select the Gaussian Process (GP) as the surrogate model, import the initial data of step 1 into the surrogate model, train the surrogate model, and when performing Bayesian optimization, first select the objective function and the acquisition function, maximize the acquisition function in the input space, and find the next experimental point; conduct experiments on the experimental point to obtain the data of the objective function, add the new experimental data to the training set, and update the surrogate model; repeat the above iterative process until the preset conditions are met.

[0016] Preferably, the objective function includes yield, purity, and by-product generation amount.

[0017] Preferably, the acquisition function is Expected Improvement (EI) or Upper Confidence Bound (UCB).

[0018] Beneficial effects

[0019] Compared with the prior art, the method for optimizing the synthesis of aryl porphyrin compounds in the present invention uses a Bayesian optimization model to systematically evaluate reaction conditions, which can maximize experimental efficiency. Specifically, the Bayesian model can provide scientific decision-making support under high uncertainty by establishing a probability relationship between reaction variables and the objective function. By introducing an artificial intelligence-assisted optimization method, not only can the synthesis efficiency of tetraphenylporphyrin be improved, but also the number of experimental tests can be minimized, thereby reducing the consumption of chemical reagents and alleviating environmental pressure. The present invention provides a new solution for the synthesis of aryl porphyrin compounds, laying a more solid foundation for research and application in related fields.

[0020] Compared with Adler's method, while increasing the yield, it minimizes the generation of by-products, further improves the purity of the final product, and improves the repeatability of the experiment, thereby enhancing the possibility of industrial production of tetraphenylporphyrin. Traditional experimental designs often rely on rules of thumb and may not comprehensively cover all possible reaction conditions, resulting in limitations in optimization results. The present invention precisely aims to solve various problems existing in the current synthesis process. By introducing machine learning and combining the advantages of artificial intelligence technology, it optimizes reaction conditions, aiming to achieve efficient and green synthesis of tetraphenylporphyrin. It promotes the development of the entire chemical synthesis field towards intelligence and greenness. Brief description of the drawings

[0021] Figure 1 It is a flowchart of the method for optimizing the synthesis of aryl porphyrin compounds in the present invention.

[0022] Figure 2 It is a variance diagram of each group between the results obtained by iterating different times through machine learning in the present invention and the optimal results.

[0023] Figure 3 1H NMR spectrum of 5,10,15,20-tetraphenylporphyrin prepared in Example 4 of the present invention.

[0024] Figure 4 1H NMR spectrum of 5,10,15,20-tetrakis(4-methylphenyl)porphyrin prepared in Example 5 of the present invention.

[0025] Figure 5 1H NMR spectrum of 5,10,15,20-tetrakis(4-methoxyphenyl)porphyrin prepared in Example 6 of the present invention.

[0026] Figure 6 1H NMR spectrum of 5,10,15,20-tetrakis(4-cyanophenyl)porphyrin prepared in Example 7 of the present invention.

[0027] Figure 7 1H NMR spectrum of 5,10,15,20-tetrakis(3,5-dimethylphenyl)porphyrin prepared in Example 8 of the present invention.

[0028] Figure 8 1H NMR spectrum of 5,10,15,20-tetrakis(4-hydroxyphenyl)porphyrin prepared in Example 9 of the present invention.

[0029] Figure 9 1H NMR spectrum of 5,10,15,20-tetrakis(4-chlorophenyl)porphyrin prepared in Example 10 of the present invention.

[0030] Figure 10 1H NMR spectrum of 5,10,15,20-tetrakis(4-bromophenyl)porphyrin prepared in Example 11 of the present invention.

[0031] Figure 11 1H NMR spectrum of 5,10,15,20-tetrakis(4-carboxyphenyl)porphyrin prepared in Example 12 of the present invention.

[0032] Figure 12 1H NMR spectrum of 5,10,15,20-tetrakis(4-pyridyl)porphyrin prepared in Example 13 of the present invention. Detailed implementation mode

[0033] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described 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 the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0034] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. However, the present invention is not limited to the described embodiments.

[0035] Example 1

[0036] In the synthesis of tetraphenylporphyrin, machine learning is used to assist in screening the optimal conditions. In the early stage, corresponding data needs to be collected and imported into machine learning. Through literature research and preliminary studies, the following key reaction variables are finally determined: the concentration of additives, the type and dosage of solvents, the molar ratio of reactants, the type and dosage of acidic solvents, and some reaction conditions (temperature, time, and whether the substrate is distilled). According to the key variables, a fractional factorial experimental design method is used to design experimental groups and conduct experiments accordingly, recording the experimental phenomena and results. The experimental results obtained are systematically collected and sorted, including the following key information, namely the initial database:

[0037] Reaction variables: including the concentration of additives, the type and dosage of solvents used for reactants, the molar ratio of reactants, the type and dosage of acidic solvents, and some reaction conditions, etc.

[0038] Objective functions: including yield, purity, by-product generation amount, etc.

[0039] The preset condition is a yield of 100%.

[0040] The reaction formula is as follows:

[0041]

[0042] Example 2

[0043] The Bayesian optimization algorithm is used to find the best combination of reaction conditions, thereby improving the synthesis efficiency and product quality. The optimization process is shown in Table 1.

[0044]

[0045] Before using Bayesian optimization, a surrogate model needs to be established. In the present invention, the Gaussian process (GP) is used as the surrogate model. The Gaussian process is a non-parametric statistical model that can provide a good approximation of the objective function and can calculate uncertainty. The Gaussian process model is trained through the database constructed in Example 1 so that it can accurately predict the reaction performance under untested conditions. When performing Bayesian optimization, the objective function needs to be determined first. In the present invention, the objective function is defined as the product performance index (yield, purity, etc.). Select a suitable acquisition function (such as expected improvement or upper confidence bound), optimize the acquisition function on the surrogate model, maximize the acquisition function in the input space, and find the next experimental point x next , and at the same time at xnext Experiments were conducted at [specific location] to obtain the corresponding reaction data f(x next ), and it was added to the data warehouse to update the surrogate model. The Bayesian optimization algorithm used in the present invention refers to the article by Abigail G. Doyle. The above iterative process was repeated until the preset conditions were met.

[0046] Example 3

[0047] During the optimization process using the optimization algorithm, first, verification experiments were conducted based on the results output by the optimization algorithm, and then the model was continuously updated according to the experimental results, and the reaction conditions were dynamically adjusted. Through continuous verification experiment feedback and the Bayesian optimization process, the optimal reaction conditions were gradually approached. When optimizing, different combinations of each reaction variable were divided into four groups, and the details of the groups are shown in Tables 1 - 4. Different numbers of iterations were carried out, and finally, the predicted results were analyzed with the actual results, and the results are as shown in the appendix Figure 3 . According to this figure, it can be obtained in which round the optimal conditions for each group can be obtained.

[0048] The steps of the verification experiment are as follows: According to the summary of the optimal results output by machine learning, verification experiments were carried out. Propionic acid and p-toluenesulfonic acid (since some solid aldehydes are not soluble in toluene and were directly added to propionic acid) were added to a three-necked flask according to the simulated optimal amount and heated to the optimal temperature for reaction. Then, pyrrole and benzaldehyde were mixed and dissolved in toluene, transferred to a dropping funnel, and dropped into the propionic acid mixed solution, and the dropping was carried out according to the optimal reaction time. After the reaction ended, it was cooled and filtered by suction, and the obtained crude product was washed with ethyl acetate. After washing, it was dried in an oven overnight.

[0049] Table 1

[0050]

[0051]

[0052] Among them, a The reaction results were predicted by AI. b Performance indicators obtained through experimental verification. n / a c Not applicable (randomly selected reaction). Other reaction conditions were: the molar ratio of reactants benzaldehyde to pyrrole was 1:1, both were 200 mmol, the solvent volume was 50 mL, the acidic solvent volume was 150 mL, the additive was p-toluenesulfonic acid, the reaction temperature was 140 °C, the reaction time was 120 min, and the substrate was not distilled.

[0053] Table 2

[0054]

[0055]

[0056] Among them, a The reaction results are predicted by AI. b Performance indicators obtained through experimental verification. n / a c Not applicable (randomly selected reaction). Other reaction conditions are as follows: the molar ratio of reactant benzaldehyde to pyrrole is 1:1, both are 200 mmol, the acidic solvent is propionic acid, the additive is p-toluenesulfonic acid, the amount of the additive is 5 mol%, the reaction temperature is 140 °C, the reaction time is 120 min, and the substrate is not distilled.

[0057] Table 3

[0058]

[0059] Among them, a The reaction results are predicted by AI. b Performance indicators obtained through experiments. n / a c Not applicable (randomly selected reaction). Other reaction conditions are as follows: the solvent is toluene, the amount is 50 mL, the acidic solvent is propionic acid, the amount is 150 mL, the additive is p-toluenesulfonic acid, the reaction temperature is 140 °C, the reaction time is 120 min, and the substrate is not distilled.

[0060] Table 4

[0061]

[0062] Among them, a The reaction results are predicted by AI. b Performance indicators obtained through experiments. n / a c Not applicable (randomly selected reaction). Y - Substrate distilled. N – Substrate not distilled. Other reaction conditions are as follows: the molar ratio of reactant benzaldehyde to pyrrole is 1:1, which is 200 mmol, the solvent is toluene, the amount is 50 mL, the acidic solvent is propionic acid, the amount is 150 mL, and the additive is p-toluenesulfonic acid, the amount is 5 mol%.

[0063] Table 1-2 shows partial prediction results and experimental verification results obtained by iterating the optimization algorithm 5 times in the examples of the present invention.

[0064] Example 4

[0065] Under the optimal conditions obtained in Example 3, specific experimental verification was carried out, and the operation is as follows:

[0066] In a 10000 mL three-necked flask, 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added, and the mixture was heated to reflux at 140 °C. 425 mL of 4 mol of benzaldehyde and 280 mL of 4 mol of pyrrole were dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. The mixture was refluxed for 2 hours, cooled, filtered by suction, and the purple solid was washed twice with ethyl acetate and transferred to an oven and dried overnight at 120 °C. 393 g of 5,10,15,20-tetraphenylporphyrin was obtained with a yield of 64%. 1 H NMR(CDCl 3 , 400 MHz) as attached Figure 3 .

[0067] Example 5

[0068] In a 10000 mL three-necked flask, 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added, and the mixture was heated to reflux at 140 °C. 480 mL of 4 mol of 4-methylbenzaldehyde and 280 mL of 4 mol of pyrrole were dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. The mixture was refluxed for 2 hours, cooled, filtered by suction, and the purple solid was washed twice with methanol and transferred to an oven and dried overnight at 120 °C. 559 g of 5,10,15,20-tetrakis(4-methylphenyl)porphyrin was obtained with a yield of 83%. 1 H NMR(CDCl 3 , 400 MHz) as attached Figure 4 .

[0069] Example 6

[0070] In a 10000 mL three-necked flask, 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added, and the mixture was heated to reflux at 140 °C. 545 mL of 4 mol of 4-methoxybenzaldehyde and 280 mL of 4 mol of pyrrole were dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. The mixture was refluxed for 2 hours, cooled, filtered by suction, and the purple solid was washed twice with methanol and transferred to an oven and dried overnight at 120 °C. 442 g of 5,10,15,20-tetrakis(4-methoxyphenyl)porphyrin was obtained with a yield of 65%. 1 H NMR(CDCl 3 , 400 MHz) as attached Figure 5 .

[0071] Example 7

[0072] In a 10000 mL three-necked flask, 4 mol (537 g) of 4-cyanobenzaldehyde was taken and 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added. It was heated to reflux at 140 °C. 4 mol (280 mL) of pyrrole was dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. It was refluxed for 2 hours, cooled, and filtered by suction. The black solid was washed twice with methanol, ethyl acetate, and tetrahydrofuran respectively to become a purple solid, which was transferred to an oven and dried overnight at 120 °C. 336 g of 5,10,15,20-tetrakis(4-cyanophenyl)porphyrin was obtained, with a yield of 46%. 1 H NMR(CDCl 3 , 400 MHz) as attached Figure 6 。

[0073] Example 8

[0074] In a 10000 mL three-necked flask, 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added. It was heated to reflux at 140 °C. 4 mol (540 mL) of 3,5-dimethylbenzaldehyde and 4 mol (280 mL) of pyrrole were dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. It was refluxed for 2 hours, cooled, and filtered by suction. The remaining purple solid was washed twice with ethyl acetate and methanol respectively, transferred to an oven, and dried overnight at 120 °C. 428 g of 5,10,15,20-tetrakis(3,5-dimethylphenyl)porphyrin was obtained, with a yield of 60%. 1 H NMR(CDCl 3 , 400 MHz) as attached Figure 7 。

[0075] Example 9

[0076] In a 10000 mL three-necked flask, 4 mol (490 g) of 4-hydroxybenzaldehyde was taken and 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added. It was heated to reflux at 140 °C. 4 mol (280 mL) of pyrrole was dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. It was refluxed for 2 hours, cooled, 400 mL of ethanol was added and heated to reflux for 30 minutes, cooled to room temperature, and filtered by suction to obtain a black solid. The solid product was repeatedly washed with a 1:1 (v / v) ethanol and propionic acid solution and hot water until the washing liquid was no longer dark in color. The purple solid was washed twice with dichloromethane, transferred to an oven, and dried overnight at 120 °C. 340 g of 5,10,15,20-tetrakis(4-hydroxyphenyl)porphyrin was obtained, with a yield of 50%. 1 HNMR(DMSO-d6, 400 MHz) as attached Figure 8 。

[0077] Example 10

[0078] In a 10000 mL three-necked flask, 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added, and the mixture was heated to reflux at 140 °C. 4 mol (563 mL) of 4-chlorobenzaldehyde and 4 mol (280 mL) of pyrrole were dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. The mixture was refluxed for 2 hours, cooled, filtered by suction, and the purple solid was washed twice with methanol, transferred to an oven, and dried overnight at 120 °C. 476 g of 5,10,15,20-tetrakis(4-chlorophenyl)porphyrin was obtained, with a yield of 63%. 1 H NMR(CDCl 3 , 400 MHz) is as attached Figure 9 .

[0079] Example 11

[0080] In a 10000 mL three-necked flask, 740 g of 4-bromobenzaldehyde, 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added, and the mixture was heated to reflux at 140 °C. 4 mol (280 mL) of pyrrole was dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. The mixture was refluxed for 2 hours, cooled, filtered by suction, and the purple solid was washed twice with methanol, transferred to an oven, and dried overnight at 120 °C. 468 g of 5,10,15,20-tetrakis(4-bromophenyl)porphyrin was obtained, with a yield of 50%. 1 H NMR(CDCl 3 , 400 MHz) is as attached Figure 10 .

[0081] Example 12

[0082] In a 10000 mL three-necked flask, 4 mol (604 g) of 4-carboxybenzaldehyde, 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added, and the mixture was heated to reflux at 140 °C. 4 mol (280 mL) of pyrrole was dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. The mixture was refluxed for 2 hours, cooled, filtered by suction, and the dark purple solid was washed three times with dichloromethane, transferred to an oven, and dried overnight at 120 °C. 185 g of 5,10,15,20-tetrakis(4-carboxyphenyl)porphyrin was obtained, with a yield of 25%. 1 H NMR(DMSO-d6, 400 MHz) is as attached Figure 11 .

[0083] Example 13

[0084] In a 10000 mL three-necked flask, 3000 mL of propionic acid and 38 g of p-toluenesulfonic acid with a catalytic amount of 5 mol% were added, and the mixture was heated to reflux at 140 °C. 429 mL of 4-pyridinecarboxaldehyde (4 mol) and 280 mL of pyrrole (4 mol) were dissolved in 1000 mL of toluene, transferred to a 2000 mL dropping funnel, and then added dropwise to the propionic acid solution over 50 minutes. The mixture was refluxed for 2 hours, cooled, 200 mL of N,N-dimethylformamide was added, and the solvent was removed by distillation under reduced pressure. The remaining solid was washed twice with methanol and once with a small amount of tetrahydrofuran to obtain a purple solid, which was transferred to an oven and dried overnight at 120 °C. 310 g of 5,10,15,20-tetrakis(4-pyridyl)porphyrin was obtained with a yield of 50%. 1 HNMR(CDCl 3 ,400MHz) as attached Figure 12 .

[0085] As can be seen from the above examples, the present invention uses a Bayesian optimization model to systematically evaluate reaction conditions, weigh out appropriate acids (such as propionic acid and p-toluenesulfonic acid) in optimal amounts and add them to a three-necked flask, and then heat the mixture to an ideal reflux temperature. At the same time, pyrrole and the corresponding aldehyde (such as benzaldehyde or other aromatic aldehydes, depending on the desired porphyrin derivative) are dissolved in a solvent, transferred to a dropping funnel, and the aldehyde-pyrrole mixture is added dropwise to the acid mixture within an optimized reaction time, heated, cooled after the reaction is completed, and the crude product solid is obtained by suction filtration using a Buchner funnel. Finally, the solid product is thoroughly washed with the corresponding solvent to remove residual reaction solvents and a small amount of impurities, and then dried in an oven overnight. The experimental efficiency is maximized.

[0086] In summary, according to the optimized reaction variables provided by the model of the Bayesian optimization algorithm of the present invention, by establishing a probability relationship between the reaction variables and the objective function, scientific decision-making support can be provided under high uncertainty. By introducing an optimization method assisted by artificial intelligence, not only can the synthesis efficiency of tetraphenylporphyrin be improved, but also the number of experimental tests can be minimized, thereby reducing the consumption of chemical reagents and alleviating environmental pressure. A new solution is provided for the synthesis of arylporphyrin compounds, laying a more solid foundation for research and applications in related fields.

[0087] The above schematically describes the present invention and its implementation manners. This description is not restrictive, and what is shown in the drawings is only one of the implementation manners of the present invention. The actual structure is not limited thereto. Therefore, if those of ordinary skill in the art are inspired by it and design, without creative efforts, structural manners and embodiments similar to the technical solution without departing from the purpose of the present invention, they shall fall within the protection scope of the present invention.

Claims

1. A method for optimizing the synthesis of aryl porphyrin compounds, characterized in that: The following steps are involved: Step 1, collecting initial data, wherein the initial data includes the molar ratio between reactants, the concentration of additives, the type and amount of solvents used for the reactants, the type and amount of acidic solvents, the reaction temperature, the reaction time, and distillation; Step 2: Input the initial data into the agent model and train it using the Bayesian optimization algorithm to obtain the best combination of reaction conditions; Step 3, according to the optimal reaction condition combination of step 2, a synthesis reaction is carried out according to the specific type of the aryl porphyrin compound to obtain the aryl porphyrin compound.

2. A method for optimizing the synthesis of aryl porphyrin compounds according to claim 1, characterized in that: The reaction parameters described in step 1 are derived from data obtained from literature research and routine experiments.

3. A method for optimizing the synthesis of aryl porphyrin compounds according to claim 1, characterized in that: The specific steps of training in step 2 are as follows: select Gaussian Process (GP) as the proxy model, import the initial data of step 1 into the proxy model, train the proxy model, and then perform Bayesian optimization. First, select the objective function and acquisition function, maximize the acquisition function in the input space, and find the next experimental point; take the experimental point out for testing, obtain the data of the objective function, add the new experimental data to the training set, and update the proxy model; repeat the above iterative process until the preset conditions are met.

4. A method for optimizing the synthesis of aryl porphyrin compounds according to claim 1, characterized in that: The objective function includes yield, purity, and by-product generation.

5. A method for optimizing the synthesis of aryl porphyrin compounds according to claim 1, characterized in that: The acquisition function is the expected improvement or upper confidence limit.

Images