Cellulose green solvent screening method, device and equipment based on machine learning

Abstract

The embodiment of the invention provides a cellulose green solvent screening method, device and equipment based on machine learning, and the method comprises the steps: taking a Kamlett-Tft solvation parameter as a screening basis of the dissolving performance of cellulose, constructing and optimizing a machine learning model through a preset data driving method, and obtaining an optimized machine learning model; sMILES character string conversion is carried out on a brand new ionic liquid chemical structural formula composed of a preset Lewis acid and a preset Lewis base, and molecular descriptors are calculated and serve as features to be input into the optimized machine learning model; the optimized machine learning model is used for predicting Kamlett-Taft solvation parameters beta and beta-alpha, and the optimal combination of the ionic liquid is obtained. Therefore, rapid and efficient screening of the ionic liquid with excellent cellulose dissolving performance is achieved, compared with a traditional trial and error method, the development period is greatly shortened, the cost is reduced, and the developed ionic liquid also has certain universality.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cellulose solvent development, and relates to a cellulose green solvent screening method, device and equipment based on machine learning. BACKGROUND

[0002] Cellulose is the most widely distributed and most abundant natural polymer in nature, and is considered one of the most potential materials to replace fossil resources. However, natural cellulose is a linear macromolecule composed of D-glucopyranose groups connected by β-1,4 glycosidic bonds, and contains a large number of intramolecular and intermolecular hydrogen bonds in its structure, which are packed into a dense aggregate network structure, making cellulose insoluble in most conventional solvents, and extremely difficult to modify and process, which seriously hinders the high-value application of cellulose. Developing cellulose solvents with excellent solubility is of great significance to promote the practical application of cellulose.

[0003] At present, researchers have developed a variety of cellulose solvents, mainly divided into three categories: organic solvents, alkali / urea systems, and ionic liquids. The organic solvents, such as N-methylmorpholine-N-oxide (NMMO) and dimethyl sulfoxide (DMSO), were the first solvents developed for dissolving cellulose. Lyocell fibers produced by NMMO solvent system have been industrialized, but there are also the following shortcomings: cellulose is prone to oxidative degradation during dissolution; NMMO has poor thermal stability and is easily decomposed at high temperatures (>120℃); the solvent recovery technology of NMMO is still immature, and the economic benefit is low. In the early 20th century, the alkali / urea aqueous solution system developed by Academician Zhang Linan can quickly dissolve cellulose at low temperature, but there are still many problems in its large-scale application, such as the need for freezing before dissolution, the ability to only dissolve cellulose with low degree of polymerization (DP <600), and the high degree of cellulose degradation by the solvent. Ionic liquids (ILs) are usually organic salts composed of anions and cations, which are in liquid state at room temperature. Compared with other cellulose solvents, ILs have excellent solubility, stable chemical and thermal properties, and are therefore considered as green and environmentally friendly solvents for sustainable processing. This is because ionic liquids can form stable hydrogen bond interactions between hydrogen bonds and the hydroxyl groups of cellulose, effectively breaking the hydrogen bond constraints between cellulose molecules, allowing the cellulose molecules to depolymerize and dissolve. However, the main challenge of current research is the large number of ionic liquids, of which there are thousands of ionic liquids developed. The traditional experimental trial-and-error method of screening mechanism is costly and time-consuming, and the developed ionic liquids are not universal.

[0004] Therefore, there is still a need to find a fast and efficient method to widely screen ionic liquids with high cellulose solubility, which is of great significance to improve the processing performance of cellulose and promote the wide application of cellulose. SUMMARY

[0005] Based on the problems in the related art, the embodiments of the present application provide a cellulose green solvent screening method, device and equipment based on machine learning.

[0006] The technical solution of the embodiments of the present application is as follows:

[0007] In a first aspect, the embodiments of the present application provide a cellulose green solvent screening method based on machine learning, which comprises:

[0008] Obtaining Kamlet-Taft solvation parameters of ionic liquids from multiple data sources;

[0009] Using the Kamlet-Taft solvation parameters as the screening basis for cellulose solubility, and constructing and optimizing a machine learning model through a preset data-driven method to obtain an optimized machine learning model;

[0010] Converting a new ionic liquid chemical structure formula composed of a preset Lewis acid and a Lewis base into a SMILES string, calculating molecular descriptors and inputting them as features into the optimized machine learning model;

[0011] The optimized machine learning model predicts the Kamlet-Taft solvation parameters β and β-α to obtain an optimal ionic liquid combination; the optimal ionic liquid combination represents the most optimal β and β-α prediction values.

[0012] In a second aspect, the embodiments of the present application provide a cellulose green solvent screening device based on machine learning, which comprises:

[0013] An acquisition module for obtaining Kamlet-Taft solvation parameters of ionic liquids from multiple data sources;

[0014] An optimization module for using the Kamlet-Taft solvation parameters as the screening basis for cellulose solubility, and constructing and optimizing a machine learning model through a preset data-driven method to obtain an optimized machine learning model;

[0015] An input module for converting a new ionic liquid chemical structure formula composed of a preset Lewis acid and a Lewis base into a SMILES string, calculating molecular descriptors and inputting them as features into the optimized machine learning model;

[0016] A prediction module is used to predict the Kamlet-Taft solvation parameters β and β-α using the optimized machine learning model to obtain the most preferred combination of ionic liquids; the most preferred combination of ionic liquids represents one with the best overall predicted β and β-α values.

[0017] In some embodiments, the device further comprises: a reaction module for preparing the most preferred combination of the Lewis acid and the Lewis base into an ionic liquid in a certain proportion, adding cellulose to the ionic liquid, and stirring under heating conditions until the cellulose is dissolved, the heating temperature being 40 to 110°C.

[0018] In some embodiments, the optimization is based on the Kamlet-Taft solvation parameters to construct a database; the ionic liquid anion and cation structures in the database are converted to SMILES strings, and the Morgan Fingerprint is used as a molecular descriptor and added to the feature list to obtain an initial data set; the initial data set is optimized using normalization and outlier processing to obtain a characterized and optimized data set; the characterized and optimized data set is randomly divided into a training set and a test set according to a preset ratio; the training set is input into each initial machine learning model for training to obtain a trained machine learning model; and the hyperparameters in each of the trained machine learning models are adjusted using Bayesian hyperparameters to obtain an optimized machine learning model.

[0019] In some embodiments, the machine learning model includes but is not limited to a decision tree model, a gradient boosting tree model, a random forest model, a support vector machine model, and an extreme gradient boosting regression model.

[0020] In some embodiments, the screening criteria for the most preferred combination of ionic liquids include a β value ranging from 0.8 to 1.2 and a β-α value ranging from 0.35 to 0.8.

[0021] In some embodiments, the Lewis acid in the most preferred combination of ionic liquids includes, but is not limited to, monoprotic acids, diprotic acids, and triprotic acids;

[0022] The Lewis bases most preferably combined with the ionic liquid include, but are not limited to, imidazole, pyridine, amide, amidine, guanidine and amino groups.

[0023] In some embodiments, the cellulose has a degree of polymerization (DP) ranging from 100 to 1800.

[0024] In some embodiments, the mass fraction of dissolved cellulose is 0 to 38% (w / w).

[0025] In a third aspect, an embodiment of the present application provides a cellulose green solvent screening device based on machine learning, comprising: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement the cellulose green solvent screening method based on machine learning.

[0026] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium storing executable instructions for causing a processor to execute the executable instructions to implement the cellulose green solvent screening method based on machine learning.

[0027] The cellulose green solvent screening method based on machine learning provided by the embodiment of the present application first takes the Kamlet-Taft solvation parameter as the screening basis for the cellulose dissolution performance, and constructs and optimizes a machine learning model through a preset data-driven method to obtain an optimized machine learning model; converts a new ionic liquid chemical structure formula composed of a preset Lewis acid and Lewis base into a SMILES string, calculates molecular descriptors and inputs them as features into the optimized machine learning model; the optimized machine learning model predicts the Kamlet-Taft solvation parameters β and β-α to obtain an optimal combination of ionic liquids. In this way, the ionic liquids with excellent cellulose dissolution performance are quickly and efficiently screened, the development cycle is greatly shortened and the cost is greatly reduced compared with the traditional trial-and-error method, and the developed ionic liquids also have a certain universality. The ionic liquids screened by the method provided by the present application exhibit excellent cellulose dissolution capacity.

[0028] In addition, the method provides a leading solution for the development of renewable and degradable biomass materials, has front-end innovation and industrial transformation value, and helps the development of biomass economy and low-carbon cycle. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 is a flowchart of a cellulose green solvent screening method based on machine learning provided by an embodiment of the present application;

[0030] Figure 2 is a flowchart of a cellulose green solvent screening and preparation method based on machine learning provided by an embodiment of the present application (left); and a polarizing microscope image before cellulose dissolution and after cellulose dissolution in 20% (w / w) cellulose using the preferred ionic liquid screened by the present application (i.e., an ionic liquid composed of 2-tert-butyl-1,1,3,3-tetramethylguanidine and methoxyacetic acid) (right);

[0031] Figure 3 is a flowchart of a cellulose green solvent screening method based on machine learning provided by an embodiment of the present application; Figure 3 a) and β-α( Figure 3 b) of the Kamlet-Taft solvation parameter2 and MSE values;

[0032] Figure 4 Kamlet-Taft solvation parameter β of the test set of the optimal model provided by the embodiment of the present application Figure 4 a) and β-α Figure 4 b) comparison chart of predicted value and experimental value;

[0033] Figure 5 Kamlet-Taft solvation parameter β of the ionic liquid combined by Lewis acid A1-A10 and Lewis base C1-C10 provided by the embodiment of the present application Figure 5 a) and β-α Figure 5 b) prediction value schematic diagram;

[0034] Figure 6 conductivity (a) and viscosity (b) schematic diagram of the preferred five ionic liquids provided by the embodiment of the present application; Figure 6 Figure 6 composition structure schematic diagram of the cellulose green solvent screening device based on machine learning provided by the embodiment of the present application;

[0035] Figure 7 composition structure schematic diagram of the electronic device provided by the embodiment of the present application.

[0036] DETAILED DESCRIPTION Figure 8 In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings, and the described embodiments should not be regarded as limiting the present application. All other embodiments obtained by those skilled in the art without making creative efforts fall within the scope of protection of the present application.

[0037] In the following description, "some embodiments" are related to a subset of all possible embodiments, but it can be understood that "some embodiments" can be the same subset or different subset of all possible embodiments, and can be combined with each other without conflict. Unless otherwise defined, all technical and scientific terms used in the embodiments of the present application have the same meaning as understood by those skilled in the art to which the embodiments of the present application belong. The terms used in the embodiments of the present application are only for the purpose of describing the embodiments of the present application, and are not intended to limit the present application.

[0038] In the following description, "some embodiments" are related to a subset of all possible embodiments, but it can be understood that "some embodiments" can be the same subset or different subset of all possible embodiments, and can be combined with each other without conflict. Unless otherwise defined, all technical and scientific terms used in the embodiments of the present application have the same meaning as understood by those skilled in the art to which the embodiments of the present application belong. The terms used in the embodiments of the present application are only for the purpose of describing the embodiments of the present application, and are not intended to limit the present application.

[0039] ​The following describes an exemplary application of the cellulose green solvent screening device based on machine learning provided by the embodiments of the present application. The cellulose green solvent screening device based on machine learning provided by the embodiments of the present application can be implemented as a terminal or a server. In one implementation, the cellulose green solvent screening device based on machine learning provided by the embodiments of the present application can be implemented as various types of terminals such as a notebook computer, a tablet computer, a desktop computer, a mobile device, and the like. In another implementation, the cellulose green solvent screening device based on machine learning provided by the embodiments of the present application can also be implemented as a server. The server can be a standalone physical server, a server cluster composed of multiple physical servers, or a distributed system. The server can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms. The terminal and the server can be directly or indirectly connected through wired or wireless communication, which is not limited in the embodiments of the present application. In the following, an exemplary application of the cellulose green solvent screening device based on machine learning implemented as a server will be described.

[0040] The embodiments of the present application provide a cellulose green solvent screening method based on machine learning. Referring to Figure 1 , Figure 1 is a flowchart of a cellulose green solvent screening method based on machine learning provided by the embodiments of the present application. The steps shown in Figure 1 will be described.

[0041] In step S110, Kamlet-Taft solvation parameters of ionic liquids are obtained from multiple data sources.

[0042] In some embodiments, the Kamlet-Taft solvation parameters are a three-dimensional parameter system for quantifying the interaction between solvent molecules, including: π*(polarity / polarizability): reflecting the strength of the interaction between the solvent and the solute through dipole or induced dipole interaction; α(hydrogen bond donor ability): representing the ability of the solvent to provide protons to form hydrogen bonds; β(hydrogen bond acceptor ability): representing the ability of the solvent to accept protons to form hydrogen bonds.

[0043] In the present application, β parameters and β-α parameters are used. The β parameter is used to directly associate the binding ability of ionic liquids and cellulose hydroxyl groups. The higher the β value, the stronger the solubility. The β-α parameter reflects the synergistic effect of the solvent acting as a hydrogen bond acceptor and a donor at the same time. Optimizing this value can balance the solubility efficiency and the regeneration performance.

[0044] Step S120, the Kamlet-Taft solvation parameters are taken as screening basis for cellulose dissolution performance, and a machine learning model is constructed and optimized by a preset data-driven method to obtain an optimized machine learning model.

[0045] Step S130, a new ionic liquid chemical structure formula composed of a preset Lewis acid and Lewis base is converted into a SMILES string, molecular descriptors are calculated, and the molecular descriptors are taken as features and input into the optimized machine learning model.

[0046] In some embodiments, the SMILES string conversion refers to a process of converting a molecular structure into a simplified linear text representation (SMILES) or vice versa. In the present application, the anion and cation structures of ionic liquids are digitized to facilitate the calculation of molecular descriptors.

[0047] In some embodiments, the molecular descriptor is an index for quantifying a molecular structure into a numerical feature, including: topological descriptors, such as Morgan Fingerprint, encoding the connection mode of atoms in a molecule; physicochemical descriptors, such as molecular weight, LogP, and the number of hydrogen bonds.

[0048] In the present application, the Morgan Fingerprint and the Kamlet-Taft parameters are combined to construct a multi-dimensional feature space, thereby improving the prediction accuracy of the model.

[0049] Step S140, the optimized machine learning model predicts the Kamlet-Taft solvation parameters β and β-α to obtain an optimal ionic liquid combination; the optimal ionic liquid combination represents a combination with the most optimal β and β-α prediction values.

[0050] In some embodiments, the ionic liquid composed of a Lewis acid and a Lewis base has properties that can be regulated by the acid-base ratio. In the present application, the Kamlet-Taft parameters of the new acid-base combination can be predicted by theoretical calculation.

[0051] The cellulose green solvent screening method based on machine learning provided by the embodiment of the present application firstly takes the Kamlet-Taft solvation parameter as the screening basis of the cellulose dissolution performance, and constructs and optimizes a machine learning model through a preset data-driven method to obtain an optimized machine learning model; a new ionic liquid chemical structural formula composed of a preset Lewis acid and Lewis base is subjected to SMILES string conversion, molecular descriptors are calculated and input into the optimized machine learning model as features; the optimized machine learning model is used to predict the Kamlet-Taft solvation parameters β and β-α to obtain an optimal ionic liquid combination. In this way, the ionic liquid with excellent cellulose dissolution performance is quickly and efficiently screened, and compared with the traditional trial-and-error method, the development cycle is greatly shortened and the cost is reduced, and the developed ionic liquid also has certain universality. The ionic liquid screened by the method provided by the present application exhibits excellent cellulose dissolution capacity. In addition, the method provides a leading solution for the development of renewable and degradable biomass materials, has front-end innovation and industrial transformation value, and helps the development of biomass economy and low-carbon cycle.

[0052] In some embodiments, the method further comprises: preparing the optimal ionic liquid combination into an ionic liquid according to a certain proportion, adding cellulose into the ionic liquid, and stirring under heating conditions until the cellulose is dissolved, and the heating temperature is 40-110 DEG C.

[0053] In some embodiments, the above step S120 can be realized by the following steps S121-S126:

[0054] Step S121, constructing a database based on the Kamlet-Taft solvation parameter.

[0055] Step S122, performing SMILES string conversion on the anion and cation structures of the ionic liquid in the database, taking Morgan Fingerprint as a molecular descriptor and adding it to a feature list to obtain an initial data set.

[0056] Step S123, optimizing the initial data set by using normalization processing and outlier processing to obtain a feature and optimized data set.

[0057] Step S124, randomly dividing the feature and optimized data set into a training set and a test set according to a preset proportion.

[0058] Step S125, inputting the training set into each initial machine learning model for training to obtain a trained machine learning model.

[0059] Step S126, the super parameter in each of the trained machine learning model is adjusted by using Bayesian super parameter, and an optimized machine learning model is obtained.

[0060] In some embodiments, the machine learning model includes but is not limited to a decision tree model, a gradient boosting tree model, a random forest model, a support vector machine model and an extreme gradient boosting regression model.

[0061] In some embodiments, the screening criteria of the optimal combination of ionic liquids include that the range of β value is 0.8-1.2, and the range of β-α value is 0.35-0.8.

[0062] In some embodiments, the Lewis acid in the optimal combination of ionic liquids includes but is not limited to monobasic acid, dibasic acid and tribasic acid; and the Lewis base in the optimal combination of ionic liquids includes but is not limited to imidazole, pyridine, amide, amidino, guanidino and amino.

[0063] In some embodiments, the degree of polymerization of cellulose ranges from 100 to 1800, and the mass fraction of dissolved cellulose is 0-38% (w / w).

[0064] As a further preferred technical solution of the present application, in the process of preparing ionic liquids, the molar ratio of anions to cations ranges from 1:9 to 9:1.

[0065] In the following, an exemplary application of the embodiment of the present application in an actual application scenario will be described.

[0066] The present application provides a cellulose green solvent screening and preparation method based on machine learning, which includes machine learning model establishment and optimization, ionic liquid solvation parameter prediction and optimization, preferred ionic liquid cellulose dissolution performance exploration, and preferred ionic liquid conductivity and viscosity test.

[0067] The cellulose green solvent screening and preparation method based on machine learning provided by the present application is shown in the flow chart as Figure 2 (Left).

[0068] I. Machine learning model establishment and optimization

[0069] S1, search 262 literatures by taking "Ionic Liquid Kamlet-Taft Parameters" as a keyword through Web of science, obtain 545 groups of Kamlet-Taft solvation parameters of ionic liquids, and construct a database.

[0070] S2, using RDKit to convert the structure of the ionic liquid anion and cation in the database into a SMILES string, calculate the Morgan Fingerprint as a molecular descriptor, and add it to the feature list return.

[0071] S3, using normalization processing and outlier processing to optimize the model.

[0072] S4, the data set in step S3 is randomly divided into a training set and a test set in a ratio of 8:2.

[0073] S5, input the data of the training set in step S4 into a plurality of machine learning algorithms for modeling, and the machine learning model includes but is not limited to a decision tree model (Decision Tree, DT), a gradient boosting tree model (Gradient Boosting Tree, GBT), a random forest model (Random Forest, RF), a support vector machine model (Support Vector Machine, SVR), and an extreme gradient boosting regression model (Extreme Gradient Boosting Regression, XGBR).

[0074] As a further preferred technical solution of the present application, the machine learning model is preferably a random forest model (RF).

[0075] S6, use Bayesian hyperparameter adjustment model. In order to prevent overfitting in the hyperparameter adjustment process, cross-validation method is adopted.

[0076] S7, using mean squared error (Mean Squared Error, MSE) and decision coefficient (R 2 ) as the model evaluation index in step S4, the mean squared error MSE value is as low as possible below 0.002, and the decision coefficient R 2 is as close to 1 as possible, which is the preferred model. The R 2 and MSE values of each model are shown in Figure 3 .

[0077] S8, through the evaluation results of step S7, it is determined that the optimal model of Kamlet-Taft solvation parameters β and β-α is a random forest. The test set Kamlet-Taft solvation parameter β and β-α prediction value of the random forest model is compared with the experimental value, as shown in Figure 4 . Among them, the decision coefficients of β and β-α models are 0.93 and 0.90 respectively, and the MSE is 0.0082 and 0.018 respectively, all showing high prediction accuracy.

[0078] II. Prediction and optimization of ionic liquid solvation parameters

[0079] S9, the newly designed Lewis acid A1-A10 and Lewis base C1-C10 combined into a new ionic liquid is converted into a SMILES string, the molecular descriptor is calculated, and it is input as a feature into the preferred random forest model described in step S8, the prediction of Kamlet-Taft solvation parameters beta and beta-alpha is performed, and the prediction results are as shown in Figure 5 .

[0080] As a further preferred technical solution of the present application, the Lewis acid includes but is not limited to monobasic acid, dibasic acid and tribasic acid.

[0081] As a further preferred technical solution of the present application, the Lewis acid includes but is not limited to A1-A10: propionic acid, isopropionic acid, cyclohexanecarboxylic acid, methoxyacetic acid, ethoxyacetic acid, methoxypropionic acid, 2-hydroxypropionic acid, gluconic acid, acrylic acid, 1-butenoic acid, which can efficiently dissolve cellulose.

[0082] As a further preferred technical solution of the present application, the Lewis base includes but is not limited to C1-C10: 7-methyl-1,5,7-triazabicyclo[4.4.0]dec-5-ene, 2-tert-butyl-1,1,3,3-tetramethylguanidine, 1,1,3,3-tetramethylguanidine, 1,5-diazabicyclo[4.3.0]non-5-ene, 1,8-diazabicyclo[5.4.0]undec-7-ene, N-methylpyrrolidone, caprolactam, 3-aminopropanol, propylamine, propylenediamine, which can efficiently dissolve cellulose.

[0083] S10, as a further preferred technical solution of the present application, the screening of the optimal combination of ionic liquids is based on the beta value range of 0.8-1.2, the beta-alpha value range of 0.35-0.8, which can efficiently dissolve cellulose.

[0084] S11, according to the prediction results and preferred solutions described in steps S9 and S10, the preferred five ionic liquids are determined as C2A4 (2-tert-butyl-1,1,3,3-tetramethylguanidine methoxyacetate), C2A5 (2-tert-butyl-1,1,3,3-tetramethylguanidine ethoxyacetate), C2A6 (2-tert-butyl-1,1,3,3-tetramethylguanidine methoxypropionate), C3A4 (1,1,3,3-tetramethylguanidine methoxyacetate), C3A5 (1,1,3,3-tetramethylguanidine ethoxyacetate). These five ionic liquids have the most optimal beta and beta-alpha prediction values.

[0085] Three, the preferred ionic liquid for cellulose dissolution performance

[0086] S12, the preferred five kinds of ionic liquid C2A4, C2A5, C2A6, C3A4 and C3A5 are used to explore the cellulose dissolution performance.

[0087] The cellulose dissolution performance of the preferred ionic liquid of the application is further described below in combination with specific examples.

[0088] Example 1

[0089] Accurately weigh 3.56 g of dried cotton fiber (DP = 1069 ± 50) for use. Accurately weigh 13.70 g of 2-tert-butyl-1,1,3,3-tetramethyl guanidine and 7.21 g of methoxyacetic acid, and mix them at room temperature to prepare 2-tert-butyl-1,1,3,3-tetramethyl guanidine methoxyacetate ionic liquid (C2A4) with a molar ratio of 5:5. Add 3.56 g of dried cotton fiber, heat in a 70°C oil bath, and mechanically stir at a speed of 100 rpm. Take samples every 1 h for polarizing microscope observation. When no obvious cellulose microfibrils are observed in the polarizing microscope at a magnification of 200 times, it indicates that the cellulose is completely dissolved, thereby preparing a 17.0% (w / w) cellulose solution.

[0090] Example 2

[0091] Accurately weigh 3.31 g of dried cotton fiber (DP = 1069 ± 50) for use. Accurately weigh 12.85 g of 2-tert-butyl-1,1,3,3-tetramethyl guanidine and 7.81 g of ethoxyacetic acid, and mix them at room temperature to prepare 2-tert-butyl-1,1,3,3-tetramethyl guanidine ethoxyacetate ionic liquid (C2A5) with a molar ratio of 5:5. Add 3.31 g of dried cotton fiber, heat in a 70°C oil bath, and mechanically stir at a speed of 100 rpm. Take samples every 1 h for polarizing microscope observation. When no obvious cellulose microfibrils are observed in the polarizing microscope at a magnification of 200 times, it indicates that the cellulose is completely dissolved, thereby preparing a 16.0% (w / w) cellulose solution.

[0092] Example 3

[0093] Accurately weigh 4.41 g of the dried cotton fiber (DP = 1069 ± 50) and keep it ready for use. Accurately weigh 13.70 g of 2-tert-butyl-l,l,3,3-tetramethylguanidine and 8.33 g of methoxypropionic acid and mix them at room temperature under stirring to prepare 2-tert-butyl-l,l,3,3-tetramethylguanidine methoxypropionate ionic liquid (C2A6) in a molar ratio of 5:5. Add 4.41 g of the dried cotton fiber, heat it in a 70 °C oil bath and perform mechanical stirring at 100 rpm. Take samples every hour and observe them under a polarized light microscope. When no cellulose microfibrils are observed in the polarized light microscope at a magnification of 200x, it means that the cellulose is completely dissolved and a 20.0% (w / w) cellulose solution is prepared. The polarized light microscope image of the cellulose solution prepared in Example 3 is shown in the right image of Fig. 1. Figure 2

[0094] Example 4

[0095] Accurately weigh 2.26 g of the dried cotton fiber (DP = 1069 ± 50) and keep it ready for use. Accurately weigh 12.67 g of l,l,3,3-tetramethylguanidine and 9.91 g of methoxyacetic acid and mix them at room temperature under stirring to prepare l,l,3,3-tetramethylguanidine methoxyacetate ionic liquid (C3A4) in a molar ratio of 5:5. Add 2.26 g of the dried cotton fiber, heat it in a 70 °C oil bath and perform mechanical stirring at 100 rpm. Take samples every hour and observe them under a polarized light microscope. When no cellulose microfibrils are observed in the polarized light microscope at a magnification of 200x, it means that the cellulose is completely dissolved and an 11.0% (w / w) cellulose solution is prepared.

[0096] Example 5

[0097] Accurately weigh 2.08 g of the dried cotton fiber (DP = 1069 ± 50) and keep it ready for use. Accurately weigh 11.52 g of l,l,3,3-tetramethylguanidine and 10.41 g of ethoxyacetic acid and mix them at room temperature under stirring to prepare l,l,3,3-tetramethylguanidine ethoxyacetate ionic liquid (C3A5) in a molar ratio of 5:5. Add 2.08 g of the dried cotton fiber, heat it in a 70 °C oil bath and perform mechanical stirring at 100 rpm. Take samples every hour and observe them under a polarized light microscope. When no cellulose microfibrils are observed in the polarized light microscope at a magnification of 200x, it means that the cellulose is completely dissolved and a 9.5% (w / w) cellulose solution is prepared.

[0098] Example 6

[0099] ​Accurately weigh 3.29 g of dried cotton fiber (DP = 1069 ± 50) and keep aside. Accurately weigh 6.85 g of 2-tert-butyl-l,l,3,3-tetramethylguanidine and 16.66 g of methoxy propionic acid and mix them at room temperature to prepare 2-tert-butyl-l,l,3,3-tetramethylguanidine methoxy propionate ionic liquid (C2A6) in 2:8 molar ratio. Add 3.29 g of dried cotton fiber and heat in an oil bath at 70 °C with mechanical stirring at 100 rpm. Take samples at every 1 h interval for polarized light microscopic observation. When no cellulose microfibrils are observed in the polarized light microscope at 200x magnification, it indicates complete dissolution of cellulose and a 14.0% (w / w) cellulose solution is obtained.

[0100] Example 7

[0101] Accurately weigh 3.47 g of dried cotton fiber (DP = 1069 ± 50) and keep aside. Accurately weigh 13.70 g of 2-tert-butyl-l,l,3,3-tetramethylguanidine and 2.08 g of methoxy propionic acid and mix them at room temperature to prepare 2-tert-butyl-l,l,3,3-tetramethylguanidine methoxy propionate ionic liquid (C2A6) in 8:2 molar ratio. Add 3.47 g of dried cotton fiber and heat in an oil bath at 70 °C with mechanical stirring at 100 rpm. Take samples at every 1 h interval for polarized light microscopic observation. When no cellulose microfibrils are observed in the polarized light microscope at 200x magnification, it indicates complete dissolution of cellulose and a 22.0% (w / w) cellulose solution is obtained.

[0102] Example 8

[0103] Accurately weigh 2.97 g of dried cotton fiber (DP = 1069 ± 50) and keep aside. Accurately weigh 13.70 g of 2-tert-butyl-l,l,3,3-tetramethylguanidine and 8.33 g of methoxy propionic acid and mix them at room temperature to prepare 2-tert-butyl-l,l,3,3-tetramethylguanidine methoxy propionate ionic liquid (C2A6) in 5:5 molar ratio. Add 2.97 g of dried cotton fiber and heat in an oil bath at 55 °C with mechanical stirring at 100 rpm. Take samples at every 1 h interval for polarized light microscopic observation. When no cellulose microfibrils are observed in the polarized light microscope at 200x magnification, it indicates complete dissolution of cellulose and a 13.5% (w / w) cellulose solution is obtained.

[0104] Example 9

[0105] Accurately weigh 4.52 g of the dried cotton fiber (DP = 1069 ± 50) and keep it ready for use. Accurately weigh 13.70 g of 2-tert-butyl-l,l,3,3-tetramethylguanidine and 8.33 g of methoxy propionic acid and mix them at room temperature under stirring to prepare 2-tert-butyl-l,l,3,3-tetramethylguanidine methoxy propionate ionic liquid (C2A6) in a molar ratio of 5:5. Add 4.52 g of the dried cotton fiber and heat it in an oil bath at 80 °C with mechanical stirring at 100 rpm. Take samples at every 1 h and observe them under a polarized light microscope. When no cellulose microfibrils are observed under a polarized light microscope at a magnification of 200x, it indicates that the cellulose is completely dissolved and a 20.5% (w / w) cellulose solution is obtained.

[0106] Example 10

[0107] Accurately weigh 8.37 g of the dried cotton fiber (DP = 162 ± 50) and keep it ready for use. Accurately weigh 13.70 g of 2-tert-butyl-l,l,3,3-tetramethylguanidine and 8.33 g of methoxy propionic acid and mix them at room temperature under stirring to prepare 2-tert-butyl-l,l,3,3-tetramethylguanidine methoxy propionate ionic liquid (C2A6) in a molar ratio of 5:5. Add 8.37 g of the dried cotton fiber and heat it in an oil bath at 70 °C with mechanical stirring at 100 rpm. Take samples at every 1 h and observe them under a polarized light microscope. When no cellulose microfibrils are observed under a polarized light microscope at a magnification of 200x, it indicates that the cellulose is completely dissolved and a 38.0% (w / w) cellulose solution is obtained.

[0108] Example 11

[0109] Accurately weigh 3.19 g of the dried cotton fiber (DP = 1800 ± 50) and keep it ready for use. Accurately weigh 13.70 g of 2-tert-butyl-l,l,3,3-tetramethylguanidine and 8.33 g of methoxy propionic acid and mix them at room temperature under stirring to prepare 2-tert-butyl-l,l,3,3-tetramethylguanidine methoxy propionate ionic liquid (C2A6) in a molar ratio of 5:5. Add 3.19 g of the dried cotton fiber and heat it in an oil bath at 70 °C with mechanical stirring at 100 rpm. Take samples at every 1 h and observe them under a polarized light microscope. When no cellulose microfibrils are observed under a polarized light microscope at a magnification of 200x, it indicates that the cellulose is completely dissolved and a 14.5% (w / w) cellulose solution is obtained.

[0110] Table 1: Summary of the dissolution parameters in each of Examples 1 to 11

[0111]

[0112]

[0113] As can be seen from the examples in Table 1, the five ionic liquids screened based on machine learning all have the ability to dissolve high DP value (>1000) cellulose. Among them, 2-tert-butyl-1,1,3,3-tetramethyl guanidine methoxy propionate shows excellent cellulose solubility and can dissolve cellulose with a DP value range of 100-1800. In this solvent system, the solubility of cotton fiber with a DP value of 162 is 38% (w / w), the solubility of cotton fiber with a DP value of 1069 is 22% (w / w), and the solubility of cotton fiber with a DP value of 1800 is 14.5% (w / w).

[0114] Four, conductivity and viscosity test of preferred ionic liquid

[0115] S13, conductivities and viscosities of the preferred five ionic liquids C2A4, C2A5, C2A6, C3A4 and C3A5 are tested.

[0116] The conductivity and viscosity of the preferred ionic liquid of the present application are further described below in combination with specific examples.

[0117] It is well known that conductivity directly reflects the number and migration rate of free ions of ionic liquid, which determines the cellulose solubility of ionic liquid to some extent. According to the method of examples 1-5, C2A4, C2A5, C2A6, C3A4 and C3A5 five ionic liquids with a molar ratio of anion to cation of 5:5 are prepared, and the conductivities of the five ionic liquids are tested at a temperature range of 30-90℃, as shown in Figure 6 (a). The conductivities of the five ionic liquids all increase with the increase of temperature. The increase of temperature promotes the thermal motion of anion and cation, and enhances the migration rate of free ions. At the same temperature, the size of the five ionic liquids is C2A6>C2A4>C2A5>C3A4>C3A5, which is consistent with the trend of the solubility of ionic liquid to cellulose. Among them, C2A6 has the highest conductivity, indicating that it has the most free ions and the fastest migration rate, which is the reason for its excellent solubility performance.

[0118] The viscosity of ionic liquid is closely related to the migration rate of free ions, and the smaller the viscosity, the more conducive to the migration of free ions. According to the method of examples 1-5, C2A4, C2A5, C2A6, C3A4 and C3A5 five ionic liquids with a molar ratio of anion to cation of 5:5 are prepared, and the viscosities of the five ionic liquids are tested at a temperature range of 30-90℃, as shown in Figure 6(b) shown. The viscosity of the five ionic liquids decreases with the increase of temperature, and the increase of temperature promotes the thermal motion of anions and cations, weakens the interionic force, and leads to the decrease of viscosity. At the same temperature, the size of the five ionic liquids is C2A6 < C2A4 < C2A5 < C3A4 < C3A5, and the change trend is consistent with the solubility trend of the ionic liquid to cellulose. Among them, the viscosity of C2A6 is the lowest, indicating that the free ion migration rate is the fastest, which is the reason why it shows excellent dissolution performance.

[0119] In summary, the present application provides a cellulose green solvent screening and preparation method based on machine learning, which adopts an efficient machine learning method to model the Kamlet-Taft solvation parameters β and β-α data of ionic liquids, and successfully constructs the Kamlet-Taft solvation parameter prediction model of ionic liquids. The determination coefficients of the β and β-α models are 0.93 and 0.90, respectively, and the MSE is 0.0082 and 0.018, respectively, showing high prediction accuracy. The model is used to predict the β and β-α parameters of new ionic liquids composed of Lewis acids and Lewis bases. Among them, the types of Lewis acids include but are not limited to monobasic acid, dibasic acid and tribasic acid, and the types of Lewis bases include but are not limited to imidazole, pyridine, amide, amidino, guanidino and amino. Five preferred ionic liquids are screened from them, all of which show excellent cellulose dissolution performance. Among them, 2-tert-butyl-1,1,3,3-tetramethyl guanidino methoxy propionate shows excellent cellulose dissolution capacity and can dissolve cellulose with a DP value range of 100-1800. The present application realizes the rapid and efficient screening of ionic liquids with excellent cellulose dissolution performance, greatly shortens the development cycle and reduces the cost compared with the traditional trial-and-error method. It provides a leading solution for the development of renewable and biodegradable biomass materials, has frontier innovation and industrial transformation value, and helps the development of biomass economy and low-carbon cycle.

[0120] Figure 7 is the composition structure schematic diagram of the cellulose green solvent screening device based on machine learning provided by the embodiment of the present application, as Figure 7As shown, the cellulose green solvent screening device 700 based on machine learning comprises: an acquisition module 701, configured to acquire Kamlet-Taft solvation parameters of ionic liquids from a plurality of data sources; an optimization module 702, configured to take the Kamlet-Taft solvation parameters as screening basis of cellulose dissolution performance, and construct and optimize a machine learning model by a preset data-driven method to obtain an optimized machine learning model; an input module 703, configured to perform SMILES string conversion on a new ionic liquid chemical structure formula composed of a Lewis acid and a Lewis base, calculate molecular descriptors and input the molecular descriptors as features into the optimized machine learning model; and a prediction module 704, configured to perform prediction of Kamlet-Taft solvation parameters β and β-α by the optimized machine learning model to obtain an optimal ionic liquid combination; the optimal ionic liquid combination represents an ionic liquid combination with the most optimal β and β-α prediction values.

[0121] It should be noted that the description of the device of the embodiments of the present application is similar to the description of the above method embodiments, has similar beneficial effects as the method embodiments, and therefore will not be described again. For technical details not disclosed in the device embodiments, please refer to the description of the method embodiments of the present application for understanding.

[0122] It should be noted that in the embodiments of the present application, if the above-mentioned cellulose green solvent screening method based on machine learning is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present application or the parts that contribute to the related art can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a terminal to execute all or part of the methods described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read Only Memory), a magnetic disk or an optical disk, and various program code storage media. Therefore, the embodiments of the present application are not limited to any specific hardware and software combination.

[0123] Correspondingly, the embodiments of the present application provide an electronic device, Figure 8 is a schematic diagram of the composition structure of the electronic device provided by the embodiments of the present application, like Figure 8As shown, the electronic device 800 at least includes a processor 801 and a computer readable storage medium 802 configured to store executable instructions, where the processor 801 generally controls the overall operation of the electronic device 800. The computer readable storage medium 802 is configured to store instructions and applications executable by the processor 801, and can also cache data to be processed by the processor 801 and modules in the electronic device 800, and can be implemented by a FLASH or a Random Access Memory (RAM).

[0124] The embodiments of the present application provide a storage medium storing executable instructions, where the executable instructions, when executed by a processor, cause the processor to perform the method provided by the embodiments of the present application, for example, as shown in the method. Figure 1

[0125] In some embodiments, the storage medium can be a computer readable storage medium, for example, a Ferroelectric Memory (FRAM), a Read Only Memory (ROM), a Programmable Read Only Memory (PROM), an Erasable Programmable Read Only Memory (EPROM), an Electrically Erasable Programmable Read Only Memory (EEPROM), a flash memory, a magnetic surface memory, an optical disc, or a Compact Disk-Read Only Memory (CD-ROM), etc. The storage medium can also be various devices including one or any combination of the above storage devices.

[0126] In some embodiments, the executable instructions can be in the form of a program, software, software module, script or code, written in any form of programming language, including a compiled or interpreted language, or a declarative or procedural language, and can be deployed in any form, including being deployed as a stand-alone program or being deployed as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0127] ​By way of example, an executable instruction can be, but is not limited to, a file, a part of a file, containing high level code (e.g., a script) that can be executed by a virtual machine, interpreter, or compiler, low level code, such as machine language, machine dependent code, firmware, micro-code, hardware descriptions, or either pictures or diagrams that have associated computer readable code. The described executable instructions can be, for example but not limited to, code that publically available or developed in a proprietary environment for one specific use. An executable instruction can be deployed to be executed on one electronic device or on multiple electronic devices located at one place or distributed across multiple places and interconnected via a communication network.

[0128] The above description is only some embodiments of the present application, and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement and improvement within the spirit and scope of the present application shall fall within the protection scope of the present application.

[0129] It should be understood that the description throughout the specification can make reference to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. Therefore, the appearances of the phrase "in one embodiment" or "in an embodiment" in various places throughout the specification are not necessarily referring to the same embodiment. Furthermore, the particular features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that the sequence of the processes described above does not mean the execution order of the processes, and the execution order of the processes should be determined according to the functions and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the application. The sequence of the above embodiments of the application is only for description, and does not represent the advantages or disadvantages of the embodiments.

[0130] It should be noted that in this document, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, so that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without more limitations, an element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element. In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The above described device embodiments are only schematic, for example, the division of the units is only a logical function division, and actual implementation can be different, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0131] The above merely illustrates the embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for screening cellulose green solvents based on machine learning, characterized in that: The method comprises: Obtain Kamlet-Taft solvation parameters of ionic liquids from various data sources; The Kamlet-Taft solvation parameter is used as a screening basis for cellulose solubility performance, and a machine learning model is constructed and optimized by a preset data-driven method to obtain an optimized machine learning model; Converting the chemical structure of a new ionic liquid composed of a preset Lewis acid and Lewis base into a SMILES string, calculating the molecular descriptor and inputting it as a feature into the optimized machine learning model; The optimized machine learning model predicts the Kamlet-Taft solvation parameters β and β-α to obtain the most preferred combination of ionic liquids; the most preferred combination of ionic liquids indicates one with the best overall predicted β and β-α values.

2. The method according to claim 1, characterized in that The method further comprises: The most preferred combination of the Lewis acid and the Lewis base is prepared into an ionic liquid in a certain ratio, cellulose is added into the ionic liquid, and stirred under heating conditions until the cellulose is dissolved, and the heating temperature is 40-110°C.

3. The method according to claim 1, characterized in that The Kamlet-Taft solvation parameter is used as a screening basis for cellulose solubility performance, and a machine learning model is constructed and optimized by a preset data-driven method, including: Based on the Kamlet-Taft solvation parameters, a database is constructed; The ionic liquid anion and cation structures in the database were converted into SMILES strings, and MorganFingerprint was used as a molecular descriptor and added to the feature list to obtain an initial data set; Optimizing the initial data set by normalization and outlier processing to obtain a characterized and optimized data set; The characterized and optimized data set is randomly divided into a training set and a test set according to a preset ratio; Inputting the training set into each initial machine learning model for training to obtain a trained machine learning model; Bayesian hyperparameters are used to adjust the hyperparameters in each of the trained machine learning models to obtain an optimized machine learning model.

4. The method according to claim 1, wherein The machine learning models include but are not limited to decision tree models, gradient boosting tree models, random forest models, support vector machine models and extreme gradient boosting regression models.

5. The method according to claim 1, wherein The screening criteria for the most preferred combination of the ionic liquids include a β value range of 0.8 to 1.2 and a β-α value range of 0.35 to 0.

8.

6. The method according to claim 1, characterized in that The Lewis acid in the most preferred combination of ionic liquids includes but is not limited to monobasic acids, dibasic acids and tribasic acids; The Lewis bases most preferably combined with the ionic liquid include, but are not limited to, imidazole, pyridine, amide, amidine, guanidine and amino groups.

7. The method according to claim 2, characterized in that The degree of polymerization of the cellulose is in the range of 100 to 1800.

8. The method according to claim 2, characterized in that The mass fraction of dissolved cellulose is 0-38% (w / w).

9. A cellulose green solvent screening device based on machine learning, characterized in that: The device comprises: Acquisition module, used to obtain Kamlet-Taft solvation parameters of ionic liquids from various data sources; an optimization module for using the Kamlet-Taft solvation parameter as a screening basis for cellulose solubility performance, and constructing and optimizing a machine learning model through a preset data-driven method to obtain an optimized machine learning model; An input module, configured to convert a new ionic liquid chemical structure composed of a preset Lewis acid and a Lewis base into a SMILES string, calculate a molecular descriptor, and input the descriptor as a feature into the optimized machine learning model; A prediction module is used to predict the Kamlet-Taft solvation parameters β and β-α using the optimized machine learning model to obtain the most preferred combination of ionic liquids; the most preferred combination of ionic liquids represents one with the best overall predicted β and β-α values.

10. A cellulose green solvent screening device based on machine learning, characterized in that: include: a memory for storing executable instructions; A processor, configured to implement the machine learning-based cellulose green solvent screening method according to any one of claims 1 to 8 when executing the executable instructions stored in the memory.

Images