A prediction method for oil separation thermodynamic phase diagram based on molecular simulation

Abstract

The present invention provides a method for predicting an oil-out thermodynamic phase diagram based on molecular simulation; the prediction of the oil-out thermodynamic phase diagram is mainly divided into two parts, namely, the prediction of the oil-out solvent system and the oil-out curve; according to the molecular structure of the model substance to be predicted, the structure is analyzed and its potential dimer cluster configuration is searched, and high-precision energy and wave function calculations are performed, and the solvent system in which the oil-out phenomenon may occur is analyzed by comparing the obtained results; at the same time, molecular dynamics and molecular mechanics simulation methods are combined to establish amorphous units under specific proportions, and the viscosity, intermolecular interaction strength, surface energy and total energy of the system at different temperatures are calculated, and then the temperature at which the oil-out phenomenon occurs at a specific concentration, namely, the oil-out curve, is analyzed; the application domain of this method is clear, and it has good robustness and predictive ability, providing a reliable basis for the prediction and application of the oil-out phenomenon of organic small molecules, and has broad application prospects and guiding value.

Description

Technical Field

[0001] The invention belongs to the technical field of chemical engineering industrial crystallization, and particularly relates to a prediction method for an oil separation thermodynamic phase diagram based on molecular simulation. Background Art

[0002] During the dissolution or cooling crystallization process, after the solution reaches supersaturation, a new liquid phase appears before crystallization. This phenomenon is commonly referred to as oiling out, sometimes also referred to as liquid-liquid stratification or liquid-liquid phase separation. Oiling out generally affects product quality. Most studies have shown that oiling out can trap impurities in the crystals, reducing the purity of the final product and causing product agglomeration. Furthermore, in large-scale industrial production, the increased production costs and safety risks caused by oiling out are also significant. Oil adhering to the agitator increases stirring resistance, which in turn increases equipment energy consumption. Furthermore, oil adhering to the crystallizer walls increases fouling thermal resistance, disrupting normal heat transfer.

[0003] Current reports on the oiling-out phenomenon of small organic molecules have primarily focused on the thermodynamic properties of the oiling-out crystallization process. Many researchers primarily investigate the oiling-out phenomenon and its control by measuring phase diagrams during the oiling-out crystallization process. Studying the conditions and influencing factors of the oiling-out region through thermodynamic phase diagrams can provide direct optimization solutions for actual process design, which is of great significance for improving operating conditions, optimizing crystallization processes, and enhancing product quality. However, due to the lack of efficient and convenient measurement methods, determining the thermodynamic phase diagram of the oiling-out system is time-consuming and labor-intensive. Furthermore, the measurement process often involves the use of numerous harmful organisms and expensive organic solvents, which can easily pollute the environment.

[0004] Currently, most methods for predicting oil separation phase diagrams focus on thermodynamic model fitting, such as the NRTL model and the PC-SAFT model. One study, based on solubility data for dihydroxydiphenyl sulfone, successfully predicted oil separation in the dihydroxydiphenyl sulfone-water-acetone and dihydroxydiphenyl sulfone-water-isopropanol ternary systems using the PC-SAFT model. However, these methods all fall within the realm of macroscopic thermodynamic research, with limited accuracy and applicability. Accurately and efficiently predicting oil separation thermodynamic phase diagrams at the molecular level remains a challenge.

[0005] Therefore, exploring cheaper, more effective, faster and more universal methods for predicting oil separation thermodynamic phase diagrams remains a major and urgent challenge. Summary of the Invention

[0006] In order to overcome the shortcomings of low accuracy and narrow application range of existing oil separation thermodynamic phase diagram prediction methods based on thermodynamic models,

[0007] The present invention provides a method for predicting the thermodynamic phase diagram of oil separation based on molecular simulation. Through quantum chemical calculations based on density functional theory, wave function analysis and molecular dynamics simulation, the solvent system and oil separation curve in which the oil separation phenomenon occurs are predicted from the perspective of intermolecular interactions, which can solve the problems mentioned in the above background technology.

[0008] The technical solutions of the present invention are as follows:

[0009] (1) Determine the chemical structure of the substance to be predicted.

[0010] (2) A configuration search of the substance in the gas phase is performed to obtain a stable dimer configuration in the gas phase.

[0011] (3) An independent gradient model analysis was performed on the stable dimer configuration in the gas phase to determine the type of interaction between solute molecules in the dimer configuration. At the same time, a molecular electrostatic potential surface analysis was performed on the stable dimer configuration in the gas phase to determine the dimer configuration's ability to provide hydrogen bond donors and acceptors.

[0012] (4) The molecular polarity index value of the dimer configuration is calculated based on the results of molecular electrostatic potential surface analysis, and the solvent range is delineated by combining the results of the independent gradient model and molecular electrostatic potential surface analysis.

[0013] (5) Select a solvent within the screening range and conduct a configuration search for the substance in the solvent phase. If a stable dimer configuration is still obtained, the solvent is considered to be a solvent system that can cause oiling. If the dimer configuration is destroyed, it is necessary to change the solvent and conduct a new configuration search.

[0014] (6) After obtaining the solvent system, an amorphous unit with a specific solute concentration is constructed, and the total energy of the oil-precipitation system at a specific temperature is calculated through molecular dynamics simulation. The total energy-temperature curve can be drawn using the simulation results at all temperatures.

[0015] (7) Based on the total energy-temperature curve under a specific system, combined with the results of shear stress autocorrelation function analysis, radial distribution function analysis and surface energy calculation, the temperature at which oil separation occurs can be inferred.

[0016] (8) The oil precipitation curve of the system can be obtained by connecting the oil precipitation temperature points at different concentrations.

[0017] (9) The solvent system in which oiling occurs and the oiling curve can constitute the oiling thermodynamic phase diagram.

[0018] The substances to be tested in step (1) are pyraclostrobin, vanillin, ethyl vanillin, citric acid and benzoic acid respectively.

[0019] The dimer configuration search, geometry optimization, and energy calculation in steps (2), (3), and (5) were all pre-optimized using the B97-3c method using ORCA software and then completed with an accuracy no lower than B3LYP-D3(BJ) / ma-def2-TZVP.

[0020] The analysis of the independent gradient model and the molecular electrostatic potential surface in steps (3) and (4) is calculated using Multiwfn software by reading the energy information and wave function information in the molden file generated by ORCA software.

[0021] The solvent systems in step (4) are water, methanol, ethanol, 1-propanol, 2-propanol, 1-butanol, 2-butanol, isobutanol, 1-pentanol, 2-pentanol, isoamyl alcohol, 1,2-propylene glycol, glycerol, isopropyl ether, n-hexane, cyclohexane, n-octane, isooctane, dichloromethane, methyl acetate, ethyl acetate, n-propyl acetate, isopropyl acetate, n-butyl acetate, acetone, acetonitrile, N-methylpyrrolidone, N,N-dimethylformamide, dimethyl sulfoxide, and 1,4-dioxane.

[0022] The shear stress autocorrelation function analysis, radial distribution function analysis, and surface energy calculation in steps (6) and (7) were implemented using the Forcite module of Materials Studio software under the COMPASS force field. In the molecular dynamics and molecular mechanics simulations, an Andersen thermostat was used to control the temperature, and a Berendsen barostat was used to control the pressure.

[0023] The above method has the following beneficial effects:

[0024] a) This method can predict the oil separation phase diagrams of five substances including pyraclostrobin, and can also be applied to other organic compounds, which is more conducive to promotion and application.

[0025] b) This method uses high-precision molecular simulation and wave function analysis to examine the strength and type of interactions between molecular clusters and with solvent molecules from multiple angles, establish optimal prediction indicators, increase prediction accuracy, and enhance predictive capabilities.

[0026] c) The input file of this method only requires the molecular structure of the compound and solvent molecules, and does not require a large amount of thermodynamic data, thus avoiding the large amount of time and labor costs of data determination, as well as the use of harmful organisms and expensive organic solvents. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 : Flowchart of the oil separation thermodynamic phase diagram prediction method of the present invention;

[0028] Figure 2: The results of the independent gradient model analysis of the dimer configuration of pyraclostrobin in the present invention;

[0029] Figure 3 : Auxiliary experimental photos of the oiling-out phenomenon of vanillin, ethyl vanillin, citric acid and benzoic acid in the selected solvent system of the present invention; wherein: (a) is an auxiliary experimental photo of the oiling-out phenomenon of vanillin in water, (b) is an auxiliary experimental photo of the oiling-out phenomenon of ethyl vanillin in water, (c) is an auxiliary experimental photo of the oiling-out phenomenon of citric acid in water, and (d) is an auxiliary experimental photo of the oiling-out phenomenon of benzoic acid in water+isopropyl alcohol.

[0030] Figure 4 : Total energy-temperature curve at a specific concentration of pyraclostrobin in the present invention;

[0031] Figure 5 : A comparison diagram of the predicted and experimental results of the oil separation phase diagram of pyraclostrobin in a mixed solvent of cyclohexane and isopropanol in the present invention;

[0032] Figure 6 : Comparison diagram of the predicted and experimental results of the oil separation phase diagrams of vanillin, ethyl vanillin, citric acid and benzoic acid in their respective solvent systems capable of oil separation; wherein: (a) is a comparison diagram of the predicted and experimental results of the oil separation phase diagram of vanillin in water, (b) is a comparison diagram of the predicted and experimental results of the oil separation phase diagram of ethyl vanillin in water, (c) is a comparison diagram of the predicted and experimental results of the oil separation phase diagram of citric acid in water, and (d) is a comparison diagram of the predicted and experimental results of the oil separation phase diagram of benzoic acid in water+isopropyl alcohol. DETAILED DESCRIPTION

[0033] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.

[0034] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present application belongs.

[0035] A method for predicting an oil separation thermodynamic phase diagram based on molecular simulation comprises the following steps:

[0036] (1) According to Figure 1The process shown in the figure begins with modeling the substance molecules. Pyraclostrobin, vanillin, ethyl vanillin, citric acid, and benzoic acid are selected as the substances to be predicted. Water, methanol, ethanol, 1-propanol, 2-propanol, 1-butanol, 2-butanol, isobutanol, 1-pentanol, 2-pentanol, isopentanol, 1,2-propylene glycol, glycerol, isopropyl ether, n-hexane, cyclohexane, n-octane, isooctane, dichloromethane, methyl acetate, ethyl acetate, n-propyl acetate, isopropyl acetate, n-butyl acetate, acetone, acetonitrile, N-methylpyrrolidone, N,N-dimethylformamide, dimethyl sulfoxide, and 1,4-dioxane are selected as candidate solvent systems, and the molecules are modeled.

[0037] (2) According to Figure 1 Following the process shown, a gas-phase configuration search was performed on the substance. Using ORCA software, these various dimer configurations were pre-optimized using the B97-3c method. The single-point energies of each dimer configuration were then calculated with an accuracy no less than that of B3LYP-D3(BJ) / ma-def2-TZVP, and the ranking was performed, ultimately yielding a group of stable dimer configurations in the gas phase.

[0038] (3) According to Figure 1 The process shown is to perform surface electrostatic potential analysis and independent gradient model analysis on the dimer configuration that is stable in the gas phase. Based on Multiwfn software, by reading the energy information and wave function information in the molden file calculated by ORCA software, an independent gradient model analysis is performed on the dimer configuration that is stable in the gas phase to determine the type of interaction between the solute molecules in the dimer configuration. The area surrounded by the gradient isosurface of non-covalent interactions indicates the area where weak interactions exist, and the color mapped by the corresponding scatter plot indicates the type of interaction (blue represents hydrogen bonding, green represents dispersion, and red represents steric hindrance). At the same time, a molecular electrostatic potential surface analysis is performed on the dimer configuration that is stable in the gas phase to determine whether the dimer configuration has changed in its ability to provide hydrogen bond donors and acceptors compared to the original individual molecules. Changes in the donor and acceptor capacities will lead to changes in the ability to bind to a specific solvent.

[0039] (4) According to Figure 1 The process shown in the figure calculates the molecular polarity index of the dimer configuration based on the results of molecular electrostatic potential surface analysis. The magnitude of the molecular polarity index reflects the magnitude of the polarity of the dimer configuration. The solvent range is delineated by combining the results of the independent gradient model and the molecular electrostatic potential surface analysis. For example, in the dimer configuration of pyraclostrobin, there are large green areas between the solute molecules, indicating that the electron density in the corresponding weak interaction area is very low, and therefore it is a π...π stacking interaction dominated by dispersion (such as Figure 2), and the results of molecular electrostatic potential surface analysis show that its positive extreme value is located at the methoxy-OCH3 position, and the negative extreme value is located at the ester carbonyl C=O position. Compared with the single pyraclostrobin molecule, the hydrogen bond donor ability of the dimer configuration is greatly reduced, but it still maintains a high level of hydrogen bond acceptor ability. Therefore, the selected solvent system can only maintain the original dimer configuration if it meets the two conditions of having a weak hydrogen bond donor ability and a large difference in polarity from the dimer configuration. Therefore, it can be judged that solvents with a small molecular polarity index and weak hydrogen bond donor ability (n-hexane, isooctane, cyclohexane, n-octane, isopropyl ether, 2-pentanol, isopentanol and isobutanol) will not affect the dimer configuration of the solute, and oiling may occur.

[0040] (5) According to Figure 1 The process shown above selects solvents within the screening range and conducts a solvent-phase configuration search for the substance. Using ORCA software, these various dimer configurations are pre-optimized using the B97-3c method and the solvation model. The single-point energies of each dimer configuration are then calculated under the solvation model with an accuracy no less than that of B3LYP-D3(BJ) / ma-def2-TZVP, and the resulting configurations are ranked. Finally, a list of stable dimer configurations under a specific solvent environment is obtained. If a stable dimer configuration is still obtained, the solvent is considered a solvent system that can undergo oiling out. If the dimer configuration is destroyed, the solvent needs to be changed and the configuration search repeated. For example, configuration searches in n-hexane, isooctane, cyclohexane, n-octane, isopropyl ether, 2-pentanol, isopentanol, and isobutanol indicate that pyraclostrobin still exists as a dimer. Therefore, these solvent systems are considered solvent systems that can undergo oiling out. The results of the auxiliary experiments showed that all the above solvent systems except n-octane showed oil separation, which proved the accuracy of the method. Similarly, in the case of vanillin, ethyl vanillin, benzoic acid and citric acid, the above method was used to successfully predict the corresponding solvent systems that could produce oil separation. Figure 3 ((a) Vanillin (water), (b) Ethyl vanillin (water), (c) Citric acid (water) and (d) Benzoic acid (water + isopropyl alcohol)).

[0041] (6) According to Figure 1 After obtaining the solvent system, one of the solvent systems is selected to construct an amorphous unit with a specific solute concentration. Each amorphous unit contains 500 molecules, and the three sides of the amorphous unit are all greater than To eliminate the influence of periodic bond chains. Based on the Forcite module of Materials Studio software, COMPASS force field was used, and Andersen thermostat was used to control temperature, and Berendsen barostat was used to control pressure. First, irrelevant interactions were eliminated by performing 100,000 steps of geometric optimization based on molecular mechanics simulation on the periodic unit. Then, NVT (a system consisting of a fixed number of particles (N), fixed volume (V) and fixed temperature (T))-NPT (a system consisting of a fixed number of particles (N), fixed pressure (P) and fixed temperature (T))-NVT ensemble molecular dynamics simulations were carried out at different temperatures to ensure good relaxation of the system and reach equilibrium. The simulation time of each molecular dynamics process was set to 500ps and the time step was set to 1fs. At a cutoff radius of Van der Waals (vdW) interactions were calculated using the Ewald summation method with an accuracy of 0.001 kcal / mol, and electrostatic interactions were calculated using the Ewald summation method with an energy tolerance of 50,000 kcal / mol. Finally, single-point energy calculations were performed on the optimized amorphous unit to determine the total energy of the oil-precipitation system at a specific temperature. Simulation results at all temperatures, under the same composition, were used to plot a total energy-temperature curve.

[0042] (7) According to Figure 1 The process shown below is based on the total energy-temperature curve under a specific system, combined with the results of shear stress autocorrelation function analysis, radial distribution function analysis and surface energy calculation. By analyzing the total energy-temperature curve under a specific concentration, the temperature at which oil precipitation occurs can be inferred. For example, for pyraclostrobin in a mixed solvent of isopropyl alcohol and cyclohexane ( Figure 4 ), in the AB segment, the total energy of the system is primarily composed of the internal energy of the molecules, which decreases with decreasing temperature. In the BC segment, liquid-liquid phase separation occurs, the dimer configuration of the solute molecules in the system transforms, the interaction strength between the solute molecules decreases, and the surface energy of the oil droplets increases rapidly, causing the total energy of the system to gradually increase with decreasing temperature. In the CD segment, since the interaction strength between the solutes and the oiling surface energy both tend to be stable at this time, the total energy of the system is once again dominated by the internal energy of the molecules and decreases with decreasing temperature. Therefore, the temperature corresponding to the local minimum point of the total energy-temperature curve is the temperature at which oiling occurs at the current concentration.

[0043] (8) According to Figure 1 The process shown in the figure can connect the oiling-out temperature points at different concentrations to obtain the oiling-out curves of pyraclostrobin, vanillin, ethyl vanillin, citric acid and benzoic acid in the corresponding solvent systems.

[0044] (9) According to Figure 1 The process shown, the solvent system where oil separation occurs and the oil separation curve can form the oil separation thermodynamic phase diagram ( Figure 5 Pyraclostrobin (cyclohexane + isopropanol), Figure 6 Among them, (a) vanillin (water), (b) ethyl vanillin (water), (c) citric acid (water) and (d) benzoic acid (water + isopropanol)), it can be seen that the average relative deviation (ARD) value and root mean square error (RMSD) value are both small, indicating that the predicted results match the experimental results well, and the prediction results of this method are relatively accurate.

[0045] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for predicting oil separation thermodynamic phase diagram based on molecular simulation, characterized in that: The steps include: Step 1: Prediction of solvent system: The dimer configuration of the predicted substance is searched in gas phase and solvent environment to determine its stability. The solvent range is delineated by combining the results of independent gradient model analysis, molecular electrostatic potential surface analysis and molecular polarity index calculation. Specifically include: (1) Determine the chemical structure of the substance to be predicted; (2) searching for the configuration of the substance in the gas phase, and obtaining a stable dimer configuration in the gas phase; (3) An independent gradient model analysis is performed on the stable dimer configuration in the gas phase to determine the type of interaction between solute molecules in the dimer configuration; at the same time, a molecular electrostatic potential surface analysis is performed on the stable dimer configuration in the gas phase to determine the ability of the dimer configuration to provide hydrogen bond donors and acceptors at this time; (4) The molecular polarity index of the dimer configuration was calculated based on the results of the molecular electrostatic potential surface analysis, and the solvent range was delineated by combining the results of the independent gradient model and the molecular electrostatic potential surface analysis; (5) Select a solvent within the screening range and conduct a configuration search for the substance in the solvent phase; If a stable dimer configuration can still be obtained, the solvent is considered to be a solvent system where oiling occurs; if the dimer configuration is destroyed, the solvent needs to be replaced and the configuration search should be repeated; Step 2: Prediction of oiling-out curve: The selected solvent and the test substance are simulated based on the molecular dynamics method. The total energy-temperature curve of the system is plotted by analyzing the energy calculation results at different temperatures. The oiling-out temperature at different concentrations is then obtained, i.e., the oiling-out curve. Step 3: Prediction of the oil separation thermodynamic phase diagram: The solvent system in which oil separation occurs and the oil separation curve can constitute the oil separation thermodynamic phase diagram.

2. The method for predicting the oil separation thermodynamic phase diagram based on molecular simulation according to claim 1, wherein: In the step (1), the substance to be tested is any one of pyraclostrobin, vanillin, ethyl vanillin, citric acid or benzoic acid.

3. The method for predicting the oil separation thermodynamic phase diagram based on molecular simulation according to claim 1, wherein: The search for dimer configurations in steps (2), (3) and (5) is performed by pre-processing with the B97-3c method using ORCA software and then performed at an accuracy not lower than B3LYP-D3(BJ) / ma-def2-TZVP.

4. The method for predicting the oil separation thermodynamic phase diagram based on molecular simulation according to claim 1, wherein: The analysis of the independent gradient model and the molecular electrostatic potential surface in steps (3) and (4) is calculated using Multiwfn software by reading the energy information and wave function information in the molden file generated by ORCA software.

5. The method for predicting the oil separation thermodynamic phase diagram based on molecular simulation according to claim 1, wherein: The solvent system in steps (4) and (5) includes water, methanol, ethanol, 1-propanol, 2-propanol, 1-butanol, 2-butanol, isobutanol, 1-pentanol, 2-pentanol, isoamyl alcohol, 1,2-propylene glycol, glycerol, isopropyl ether, n-hexane, cyclohexane, n-octane, isooctane, dichloromethane, methyl acetate, ethyl acetate, n-propyl acetate, isopropyl acetate, n-butyl acetate, acetone, acetonitrile, N-methylpyrrolidone, N,N-dimethylformamide, dimethyl sulfoxide and 1,4-dioxane.

6. The method for predicting the oil separation thermodynamic phase diagram based on molecular simulation according to claim 1, characterized in that: The step 2 is specifically as follows: S1. After obtaining the solvent system, construct an amorphous unit with a specific solute concentration, calculate the total energy of the oil-out system at a specific temperature through molecular dynamics simulation, and use the simulation results at all temperatures to draw a total energy-temperature curve; S2. Based on the total energy-temperature curve under a specific system, combined with the results of shear stress autocorrelation function analysis, radial distribution function analysis and surface energy calculation, the temperature at which oil separation occurs can be inferred; S3. Connect the oiling-out temperature points at different concentrations to obtain the oiling-out curve of the system.

7. The method for predicting the oil separation thermodynamic phase diagram based on molecular simulation according to claim 6, characterized in that: The calculation of total energy, shear stress autocorrelation function analysis, radial distribution function analysis and surface energy calculation in steps S1 and S2 are implemented using the Forcite module of Materials Studio software under the COMPASS force field.

8. The method for predicting the oil separation thermodynamic phase diagram based on molecular simulation according to claim 6, wherein: In the molecular dynamics simulation, the Andersen thermostat is used to control the temperature and the Berendsen barostat is used to control the pressure.

Images