Directional Construction Method of Special Polyether Materials with Specific Functional Groups

Abstract

This invention discloses a method for the directional construction of special polyether materials with specific functional groups, specifically relating to the field of polymer functional materials technology. The method involves constructing a main template M1 and an auxiliary template M2, introducing polyether monomer A and functional group monomer B, and initiating a regioselective synergistic polymerization reaction under an external field to form a block segment structure G1. The orientation of monomer B is adjusted based on the tension curvature Δθ to further form a block polymer G2 with a precise sequence. Subsequently, the free radical migration rate is controlled, and a special end-capping agent E1 is added to achieve adaptive termination of the reaction, obtaining a polyether final product F with a high degree of functional group sequence order. Structural inversion verification is performed using two-dimensional NMR spectroscopy combined with a sequence recognition model to ensure the accuracy of sequence construction. This method has advantages such as strong functional group orientation, high sequence accuracy, and good structural controllability, and is applicable to fields such as oilfield adjuvants and agrochemical adjuvants.

Description

Technical Field

[0001] This invention relates to the field of polymer functional materials technology, specifically to a method for the directional construction of special polyether materials with specific functional groups. Background Technology

[0002] The synthesis of polyether materials is relatively complex, and the introduction of different functional groups requires specific reaction conditions and catalysts, resulting in a high technical threshold. This technology demands a deep understanding of the regulatory mechanisms of oil-water interfacial interactions and a precise grasp of how to transform the target molecular structure from theory to reality through chemical synthesis. Specifically, the key lies in developing one or more efficient and controllable synthetic strategies to precisely construct polyether molecular chains containing specific functional groups, chain lengths, and degrees of branching. These synthetic routes need to be able to flexibly adjust reaction conditions, such as temperature, pressure, and catalyst concentration, to achieve fine control over the polyether molecular structure. Simultaneously, the economic efficiency, environmental friendliness, and scalability of the synthesis process must be ensured to meet application requirements. Through continuous optimization and improvement of synthetic routes, special polyether materials with excellent emulsifying or demulsifying properties can be efficiently prepared.

[0003] In the construction of biomimetic channels in microfluidic chips, polyether materials are required to exhibit a gradient arrangement of specific functional groups on a spatial scale to simulate the surface energy changes of natural microstructures. However, current technologies cannot precisely control the distribution position and proportion of specific functional groups (such as carboxyl groups, mercapto groups, isocyanate groups, etc.) on the polyether backbone or branches. Typically, initiators or crosslinking agents are randomly introduced, failing to form stable, directional structures. Excessive treatment of the polyether molecular structure to enhance functional group orientation can easily lead to backbone degradation or abnormal chain length distribution, thereby compromising the material's original flexibility. Summary of the Invention

[0004] The purpose of this invention is to provide a method for the directional construction of special polyether materials with specific functional groups, so as to overcome the shortcomings of the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for the directional construction of special polyether materials with specific functional groups, comprising:

[0006] A dual-template cooperative structure M1 and M2 was constructed, where M1 is the main template with periodically arranged spatial recognition sites P1; M2 is the auxiliary template containing a response steering domain P2 that can respond to changes in the chemical potential energy of reactants.

[0007] Introduce the target polyether monomer A and the predetermined functional group monomer B into the reaction system in which M1 and M2 coexist. Under the control of an external field, initiate a regioselective co-polymerization reaction to form a block segment structure G1 with initial orientation characteristics.

[0008] Based on the molecular conformation parameters of G1, the tension curvature Δθ between molecular chain segments is monitored in real time. Through synergistic coupling with the response guidance domain P2 of M2, the connection orientation of monomer B on the main chain is automatically adjusted to form a precision block polymer G2 with a functional group sequence uniformity of not less than 92%.

[0009] Under the condition of controlling the free radical migration rate Rd, a special end-capping agent E1 is introduced to adaptively terminate the unfinished chain segment structure in G2, generating a special polyether final product F with a mean square deviation σ of functional group sequence order of no more than 10%.

[0010] The functional group-oriented construction effect of the final product F was verified by using two-dimensional nuclear magnetic resonance spectroscopy and a sequence recognition algorithm based on a machine learning training model.

[0011] Preferably, the molecular conformational parameters based on G1, which monitor the tensile bending Δθ between molecular chain segments in real time, include:

[0012] In-situ laser Raman spectroscopy was used to scan the real-time vibrational peak positions of the block chain segment structure G1 in the reaction system, and the characteristic frequency signals of chain segment tension changes were extracted.

[0013] The characteristic frequency signal is converted into tension curvature Δθ, and the tension-conformation correlation matrix is ​​constructed using the Fourier transform algorithm;

[0014] The Δθ parameter is input into the response guidance domain P2 of M2, triggering a configurational adjustment of the rotatable group in P2, thereby spatially redirecting the insertion path of the functional group monomer B.

[0015] Preferably, the Δθ parameter is input to the response steering domain P2 of M2 to trigger a configurational adjustment of the rotatable group in P2, including:

[0016] The Δθ parameter is converted into the corresponding potential signal ε, and the change in potential difference reflects the degree of conformational deviation between chain segments.

[0017] The ε value is applied to the response unit in P2, which contains a set of rotatable conformational groups linked by aryl-ether bonds, and whose conformational angles can be reversibly switched within ±45°.

[0018] Driven by ε, the rotatable group undergoes configurational rearrangement, thereby changing the local polarity distribution of P2 and guiding the target functional group monomer B to approach the active recognition site on M1 along a new spatial path.

[0019] Preferably, automatically adjusting the connection orientation of monomer B on the main chain includes:

[0020] During the polymerization reaction, the local torsional conformation and chain segment strain direction of the block chain structure G1 are collected in real time, and the three-dimensional spatial orientation data of the molecular chain segments are obtained through a multi-channel angular distribution imager.

[0021] Based on spatial orientation data, the electro-algebraic response unit embedded in the auxiliary template M2 is activated, and the polarizable groups in the response guidance domain P2 undergo reversible rotation or displacement to form a guided potential well.

[0022] By utilizing potential wells to capture B monomers at the microscale, they are connected in accordance with conformational preference when they are close to the main chain, ensuring that functional groups are inserted into the main chain skeleton at the minimum configurational stress angle.

[0023] Preferably, the effect of functional group-oriented construction on the final product F is verified by inversion, including:

[0024] Two-dimensional nuclear magnetic resonance spectra of the final product F were acquired to obtain the coupling shift signals between functional groups and their cross-peak images.

[0025] The spectral data is input into the trained sequence recognition model, which is constructed based on a combination of convolutional neural networks and recurrent neural networks, and a variety of standard block permutation samples are preset as training sets.

[0026] The probability distribution map of the functional group position of each segment is output by the sequence recognition model and compared with the target construction template to determine the sequence construction deviation.

[0027] When the similarity of the comparison results is higher than the 95% threshold, the targeted construction is considered successful; otherwise, a structural reconstruction suggestion is triggered.

[0028] Preferably, the construction of the sequence recognition model includes:

[0029] Image normalization processing is performed on two-dimensional nuclear magnetic resonance spectra;

[0030] The processed spectrogram is input into a spatial feature extraction module based on a convolutional neural network to extract local image features of the coupling relationship between functional groups;

[0031] The extracted results are further input into the sequence analysis module constructed by the recurrent neural network to identify the order of functional groups in the segmental chain;

[0032] During the training of the sequence recognition model, multiple predefined polyether block arrangement maps are used as a supervised learning training set. The model weights are optimized through error backpropagation, and finally the similarity matching results between the target product and the standard construction template are output.

[0033] Preferably, determining sequence construction bias includes:

[0034] The sequence matching mapping matrix is ​​constructed by comparing the target polyether functional group arrangement sequence output by the sequence recognition model with the standard template sequence in a one-to-one correspondence.

[0035] The average position offset Δp and the missing match rate η are calculated based on this matrix, where Δp represents the mean difference between the actual and theoretical positions of functional groups, and η represents the proportion of missed identifications or mismatches.

[0036] The deviation score is constructed by combining Δp and η. The scores are expressed using a normalized scoring method, with a maximum score of 100. A score below 90 indicates a significant construction deviation.

[0037] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0038] 1. This invention achieves highly selective directional linkage of functional group monomer B in the polyether backbone by constructing a dual-template synergistic system and introducing a real-time feedback mechanism based on tension curvature Δθ. Through external field modulation, free radical migration rate limitation, and reversible configuration adjustment, the spatial precision of the polymerization sequence is effectively improved, resulting in a highly consistent functional group arrangement sequence of the final product F, thus solving the problem of "disordered functional group embedding" in existing polyether functionalization processes.

[0039] 2. This invention is the first to combine two-dimensional nuclear magnetic resonance spectroscopy with a deep learning sequence recognition model to establish an automatic inversion verification mechanism for polyether product structures. This mechanism can quickly output functional group distribution probability maps without manual spectral interpretation, and provide structural deviation scores through quantitative indicators such as offset Δp and missing rate η. It realizes closed-loop control of the entire process from synthesis and construction to structural verification, significantly improving the controllability and industrial repeatability of material structures. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0041] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] For examples, please refer to Figure 1 As shown in this embodiment, the method for directional construction of special polyether materials with specific functional groups includes:

[0044] A dual-template cooperative structure M1 and M2 was constructed, where M1 is the main template with periodically arranged spatial recognition sites P1; M2 is the auxiliary template containing a response steering domain P2 that can respond to changes in the chemical potential energy of reactants.

[0045] Introduce the target polyether monomer A and the predetermined functional group monomer B into the reaction system in which M1 and M2 coexist. Under the control of an external field, initiate a regioselective co-polymerization reaction to form a block segment structure G1 with initial orientation characteristics.

[0046] Based on the molecular conformation parameters of G1, the tension curvature Δθ between molecular chain segments is monitored in real time. Through synergistic coupling with the response guidance domain P2 of M2, the connection orientation of monomer B on the main chain is automatically adjusted to form a precision block polymer G2 with a functional group sequence uniformity of not less than 92%.

[0047] Under the condition of controlling the free radical migration rate Rd, a special end-capping agent E2 is introduced to adaptively terminate the unfinished chain segment structure in G2, generating a special polyether final product F with a mean square deviation σ of functional group sequence order of no more than 10%.

[0048] The functional group-oriented construction effect of the final product F was verified by using two-dimensional nuclear magnetic resonance spectroscopy and a sequence recognition algorithm based on a machine learning training model.

[0049] In this invention, in order to achieve precise and directional guided construction of specific functional groups on polyether molecular chains, a dual-template structure with synergistic effects is designed and adopted, including a main template M1 and an auxiliary template M2.

[0050] The master template M1, serving as the primary localization matrix in the functional group-guided reaction, is composed of ordered polymer chains with a stable skeletal structure, with spatial recognition sites P1 periodically embedded along its main chain direction. The spatial recognition sites P1 are specifically designed based on the molecular size, electronegativity distribution, and bond energy characteristics of the target functional group monomer, and have the following two functions:

[0051] Structural recognition function: The P1 site can selectively adsorb or bind monomers with pre-defined functional groups (such as carboxyl, mercapto, isocyanate groups, etc.) through electrostatic attraction, hydrogen bonding or hydrophobic repulsion mechanisms.

[0052] Position-guided function: Due to the periodic arrangement of P1, a preliminary arrangement path in one-dimensional space can be provided before the polymerization reaction starts, which restricts the insertion position of functional group monomers on the chain and avoids random doping.

[0053] The auxiliary template M2 serves as a dynamic control component, introducing responsive structural units such as electroallergenic groups, photosensitive guiding groups, or rearrangeable π-conjugated structures. M2 contains multiple response guiding domains P2, characterized by its ability to sense minute changes in chemical potential energy (e.g., monomer concentration, charge density, or local temperature rise) in the reaction system and to "guide" the recognition of the main template M1 through intramolecular conformational rearrangement.

[0054] Specifically, P2 autonomously adjusts its configurational angle and polarity direction based on real-time monitoring of local chemical environment changes during the reaction, influencing the spatial proximity behavior of adjacent functional group monomers and making them more inclined to move closer to or further away from specific P1 sites, thereby further refining the positioning accuracy. This mechanism greatly enhances the sequence control capability of functional group insertion, making the arrangement order of functional groups in the final polyether segment highly predictable and order-preserving.

[0055] M1 and M2 can be linked by non-covalent bonds or form a composite system through physical adsorption. In practice, their efficient construction and reproduction can be achieved through liquid-phase mixing self-assembly, solution wetting / drying method or microfluidic template imprinting technology.

[0056] In the implementation of this invention, the pre-constructed master template M1 and auxiliary template M2 are placed together in the same reaction system to form a dual-template synergistic polymerization environment. The reaction system is a controllable solution phase system, and the reaction medium can be an organic solvent with moderate polarity, strong thermal stability, and good compatibility with the template structure, such as N,N-dimethylformamide (DMF), tetrahydrofuran (THF), or an ionic liquid system, to ensure the stability of the template structure and the sensitive responsiveness of intermolecular interactions.

[0057] In this reaction system, target polyether monomer A and monomer B with predetermined functional groups are introduced. Monomer A is the basic building block of the linear polyether chain segment, typically selected from monomers such as ethylene oxide, propylene oxide, and tetrahydrofuran oxide, which can form a flexible main chain structure through ring-opening polymerization. Monomer B contains at least one predetermined functional group position (such as -COOH, -SH, -NCO, etc.), and its molecular structure has conformational matching with the spatial recognition site P1 of M1. Furthermore, its polarity, electron density, and other parameters can interact and couple with the response-directing domain P2 of M2.

[0058] To achieve spatially selective control of the polymerization process, this invention introduces an external field control mechanism into the reaction system. The external field may include, but is not limited to: a temperature gradient field (ΔT): regulating the local reactive region to induce preferential polymerization of chain segments in the thermally responsive region; an electric field (E): guiding functional group B towards a specific site by regulating the dipole response of the template or monomer; and a shear field or flow field (γ): inducing the block segments to extend along the template arrangement direction through flow induction, thus pre-setting the initial configuration direction of the chain segments.

[0059] Under the combined action of the aforementioned external fields, the recognition site P1 of the main template M1 preferentially binds to the functional group monomer B, and through the response guidance domain P2 of the auxiliary template M2, it adjusts the spatial parameters such as the insertion angle and connection pitch of B in real time, thereby initiating a regionally selective synergistic polymerization reaction mechanism.

[0060] This mechanism differs significantly from traditional free polymerization or disordered insertion paths, as it pre-sets the positions of functional groups and controls the structural arrangement in the early stages of polymerization. As the polymerization reaction proceeds, the polyether backbone gradually extends, while functional group B is systematically inserted into the backbone according to the template-induced spatial path, ultimately forming an initial block chain structure G1 containing multiple oriented embedded functional groups.

[0061] The G1 structure not only possesses main chain continuity and flexible retention, but also exhibits directional arrangement of functional groups on specific chain segments, providing fundamental configurational support for the subsequent construction of polyethers with high sequence precision. Characterization by scanning electron microscopy (SEM) or small-angle X-ray scattering (SAXS) revealed that the G1 chain segment structure exhibits periodic functional group spacing characteristics at the nanoscale, with a non-uniformity error of less than 15%.

[0062] In a polymerization reaction system, as the block segment structure G1 gradually forms, differences in intramolecular tension arise between its segments due to spatial arrangement, functional group insertion, and local conformational stretching. These tension changes directly affect the bending degree, torsional configuration, and spatial entry path of subsequent monomers.

[0063] To capture this type of conformational change information, this invention employs an in-situ laser Raman spectroscopy system to perform real-time scanning of the characteristic vibrational modes of the block chain structure G1 in the reaction system. The Raman laser wavelength is preferably 532 nm, and the excitation light is focused on the micro-reaction region, recorded in conjunction with a high-sensitivity spectrometer. Peak shifts and intensity variations within the interval. The bond angle tension of the block chain structure G1 is highly correlated with the C–O–C stretching vibration or torsional mode of its main chain. When the chain segment tension changes slightly, observable shifts will appear in the corresponding spectral lines.

[0064] The system employs high-resolution sampling, acquiring spectra every 0.5 seconds, and smoothing and extracting features from the spectra using a sliding window model. It utilizes a combined principal component analysis (PCA) and support vector regression (SVR) model to extract a set of characteristic frequency signals representing changes in chain segment tension, such as frequency drift amplitude, peak width variation, and intensity differences.

[0065] The extracted frequency signal is converted into an estimated value of the tensile bending angle Δθ by the Fourier transform analysis module. After Fourier transform processing, the time-frequency domain signal is converted into a spatial response signal, and the tensile bending angle Δθ between every two adjacent chain segments in G1 is further estimated using the bending coefficient conversion formula. This angle is a geometric measure of the deviation of the molecular chain conformation from the linear structure, generally ranging from 5 to 35 degrees. A value exceeding 30 degrees can be identified as a high-stress region.

[0066] Next, using Δθ as the core variable, and combining the chain segment spacing d, the degree of local polarization μ, and the known intramolecular torsional degrees of freedom τ, a multidimensional tension-conformation correlation matrix is ​​constructed. This matrix represents the conformational change trend in three-dimensional tensor form, providing numerical basis for the subsequent template response.

[0067] In this invention, the response-directing domain P2 of the auxiliary template M2 integrates multiple electroautomerism response units, each unit comprising a set of rotatable conformational groups linked by arylene-ether bonds (–Ar–O–). Each group is capable of configurational switching within a range of ±45° under the influence of an external potential.

[0068] The parameter Δθ is converted into the corresponding potential signal ε. This conversion is based on the following relationship: assuming Δθ is greater than a set threshold. (For example, at 15°), the system determines that the conformation of this chain segment is severely deviated, and the amplitude of the corresponding generated ε signal is related to Δθ. The difference between them increases linearly as a gain factor. Let... Where k is the electrical response sensitivity coefficient, with units of . .

[0069] The ε value is input to the response control circuit of P2, triggering a rearrangement of its internal rotatable conformational groups. After the rearrangement, the local dipole orientation, polarity distribution, and potential well position of P2 are reversibly adjusted. Once the local polarity change (Δμ) in this region exceeds a preset offset threshold (e.g., 0.2 Debye), a physical guiding effect on the entry direction of functional group monomer B can be achieved.

[0070] After the configurational rearrangement is completed in the P2 response, monomer B is attracted by a small directional potential and potential well before entering the recognition domain of M1. This potential well is constructed by a non-uniform electric field formed by the rotation of the P2 group, simulating a local guiding channel, guiding monomer B to preferentially enter the conformationally allowed region when it is close to the block segment structure G1.

[0071] To ensure that the connection path minimizes configurational tension, this invention further utilizes a multi-channel angular distribution imager to acquire the spatial orientation data of the segmental chain structure G1 in real time. This device employs a light scattering three-dimensional reconstruction algorithm to non-contactly record the orientation trend of the chain segment tension principal axis from outside the reaction tank, and simultaneously calibrates it using Raman spectroscopy conformational analysis data.

[0072] Ultimately, the system forms a triaxial coupling model consisting of the main chain torsion direction, template polarity path, and B monomer migration direction. This allows the B monomer to automatically follow the conformational trend of the preceding chain segment when it approaches G1, with the insertion angle controlled within ±5° of the minimum conformational stress point (Δθ_min).

[0073] During the formation of block polymer G2, the polymer chain radicals are in a dynamic growth state, and their chain segment extension rate, direction, and reaction region are closely related to the radical migration rate Rd. If effective control is lacking, some radical segments may extend to unexpected structural sites, causing defects such as functional group misalignment, uneven density, or overlapping block sequences.

[0074] To this end, this invention precisely controls Rd by adjusting the following parameters of the reaction system: the temperature gradient range is controlled within 40–55°C to prevent overheating and excessively rapid free radical activity; the initiator concentration (such as AIBN or BPO) is set at 0.01–0.05 mol / L to ensure a moderate initial free radical concentration and prevent side chain growth; a trace amount of inhibitor (such as TEMPO or HQ) is introduced and controlled within 0.5% per million to form a migration inhibition plateau in the middle stage of the reaction; and solvent polarity is adjusted, such as by using a mixed solvent THF / DMF system, to achieve fine-tuning control of the free radical migration path. Through these methods, Rd is stabilized within a certain range. Within this range, it can ensure that the chain segment growth rate is synchronized with the end-capping process.

[0075] The special end-capping agent E2 is a reactive small molecule with high steric hindrance effect and directional capture capability, and its basic structure includes the following characteristics:

[0076] It has one or more polymerizable functional groups (such as allyl or azide).

[0077] The molecule has a polar guiding group at its end (such as — — Used to identify the location of free radicals at the end of the main chain;

[0078] Conformation sites that can be activated by external fields (such as cis-trans isomers) are embedded in the main structure to facilitate time control of the reaction termination window. E2 is added using a titration-based slow-release system, which slowly adds it to the reaction system at a fixed flow rate (0.01–0.05 mL / min) through a micropump system in the later stages of the polymerization reaction (i.e., when the G2 structure tends to saturate and Δθ fluctuations slow down), allowing it to diffuse fully to the ends of unterminated chain segments.

[0079] When the free radical terminus migrates to the reaction region, the active functional group of E2 undergoes a rapid coupling reaction with the free radical at the end of the main chain (usually a radical addition reaction or single-electron transfer-induced coupling), with a rate constant greater than [missing value]. This effectively terminates the chain segment growth process.

[0080] Meanwhile, the guiding structure within the E2 molecule allows it to preferentially interact with chain ends that have less conformational distortion and more stable chain tension, achieving adaptive selective end-capping and thus avoiding the continued growth of disordered chain segments in the sequence.

[0081] The end-capping reaction is accompanied by a small amount of heat release. With real-time monitoring of the Δθ stable region marker, the system can automatically identify the polymerization completion window and achieve a smooth transition from structural control to termination regulation.

[0082] After the termination reaction was completed, the final product F was obtained through neutralization, precipitation, solvent removal, and drying. This product possesses the following structural characteristics:

[0083] The mean square deviation σ of the functional group sequence order is no higher than 10%: The distance between adjacent functional groups and the reproducibility of their positions are determined by nuclear magnetic resonance spectroscopy (NMR) and two-dimensional spectral coupling analysis (such as 2D-COSY, NOESY). The standard deviation σ is calculated by statistically analyzing the distance between functional groups on different polyether segments. It is controlled within 10% of the average value of the entire sample sequence, indicating a high degree of sequence consistency. Main chain integrity ≥ 90%: The molecular weight distribution index (PDI) is controlled between 1.2 and 1.5, indicating that the polymerization process did not cause chain breakage or multi-branch defects. High block consistency: The block size and distribution are determined by GPC (gel permeation chromatography) and MALDI-TOF mass spectrometry to ensure that the consistency deviation of the number of repeating units in the block sequence of the polyether main chain is less than 5%.

[0084] The structure of the final product F was characterized using two-dimensional nuclear magnetic resonance (2D-NMR) to obtain spatial coupling information and positional relationships between functional groups. Preferred spectra included COSY (correlation spectrum), HSQC (heteronuclear single quantum correlation), or NOESY (nuclear Eulerian exchange spectrum), depending on the interaction mode and signal intensity between functional groups.

[0085] The concentration of the test sample was controlled at 10–20 mg / mL, and the solvent used was deuterated tetrahydrofuran (THF-d8) or deuterated chloroform (THF-d8). To ensure signal clarity, measurements were taken at room temperature (approximately 25°C) using an NMR instrument with a frequency of 400 MHz or higher. Cross-peak images acquired in each spectrum represent the coupling relationships between different functional groups.

[0086] The collected two-dimensional spectral data undergoes standardization processing via an image preprocessing module, including the following steps: removing background noise and low-signal clutter; aligning and correcting the chemical shift axes in the spectrum; enhancing peaks and sharpening edges in cross-peak images; and uniformly adjusting the image size to a fixed resolution (e.g., 256×256 pixels) to meet the model input format. After spectral preprocessing, the image data is input into a trained sequence recognition model. This model is constructed based on a combination of Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN), which respectively undertake the tasks of spectral spatial feature extraction and functional group arrangement order recognition. The CNN module includes 3 convolutional layers and 2 pooling layers, using 3×3 convolutional kernels to extract the local spatial patterns of functional group coupled feature images; the RNN module uses a bidirectional long short-term memory (Bi-LSTM) network structure to map the spatial feature sequence output by the CNN into a time series, outputting the relative arrangement order of functional groups in the polyether backbone; the final output is a functional group position probability distribution map, representing the probability distribution of the presence of specific functional groups at each position.

[0087] The model training employs a supervised learning approach. The input consists of a large number of predefined standard NMR spectra of polyether block arrangements (training set), and the output is a standard sequence of functional group arrangements. The training set comprises chemically synthesized standard samples and simulated data, while the label data is manually verified by experts to ensure high accuracy.

[0088] The model training process uses cross-entropy as the loss function and employs the Adam optimizer for weight updates. Each training round contains approximately 5000 standard spectral images, and the number of training rounds is set to 50 or until the validation set accuracy converges. After training, the model can quickly predict unknown spectral samples and output a probability distribution map of the sequence structure.

[0089] The target polyether functional group sequence output by the model will be compared one by one with the preset standard construction template sequence. The specific comparison process is as follows:

[0090] Construct a sequence matching mapping matrix, where the horizontal axis represents the functional group sites in the standard sequence, the vertical axis represents the positions in the prediction results, and the matrix elements represent the matching degree of the corresponding functional group.

[0091] Based on this matching matrix, the deviation parameters of the two core structures are calculated:

[0092] Average position offset Δp: Represents the average difference between the predicted and theoretical positions of functional groups, counted in units of carbon atoms in the main chain. For example, if a carboxyl group is predicted to be at position 8, but theoretically should be at position 6, then the offset is 2.

[0093] The missing match rate η is defined as the proportion of a functional group in the standard template that is not identified or is incorrectly identified as another functional group. For example, if 4 out of 5 standard loci are correctly identified, then η = 20%.

[0094] A normalized construction bias score is calculated based on Δp and η. The calculation method is as follows: = 100 - [Weighting factor α × Δp + Weighting factor β × η × 100]; where α and β are the penalty weights for Δp ​​and η, respectively, and it is recommended to set α = 2 and β = 1 to reflect the greater impact of positional offset on structural accuracy. Final score The full score is 100 points. If ≥ 95, the targeted construction is considered successful; if If the score is below 90, it is marked as a significant deviation in structural construction, and the system will trigger the "Structural Reconstruction Suggestion" module.

[0095] After comparison and deviation calculation are completed, the results will be output to the quality control system, forming a feedback loop. If the deviation score... If the target threshold is not reached, the system will automatically link with the front-end synthesis control platform to indicate possible error sources (such as decreased template recognition accuracy, abnormal polymerization rate, template response delay, etc.) and suggest adjusting the following parameters: polymerization reaction temperature or rate; spacing between main template recognition sites; timing of capping agent addition or free radical activity control parameters.

[0096] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for the directed construction of a specific functional group specialty polyether material, characterized by: The application relates to a method for constructing a polyether with a uniform functional group sequence. The method comprises the following steps: constructing a double-template cooperative structure M1 and M2, wherein M1 is a main template with periodically arranged space recognition sites P1; M2 is an auxiliary template containing a response guide domain P2 which can respond to the chemical potential energy change of a reactant; introducing a target polyether monomer A and a predetermined functional monomer B into a reaction system coexisting with M1 and M2, and starting a region-selective cooperative polymerization reaction under the control of an external field to form a block chain segment structure G1 with an initial directional feature; based on the molecular conformation parameters of G1, the tension bending degree Delta theta between the molecular chain links is monitored in real time, and the connection orientation of the B monomer on the main chain is automatically adjusted through the cooperative coupling with the response guide domain P2 of M2 to form a precision block polymer G2 with a functional group sequence uniformity of not less than 92%; under the condition of controlling the free radical migration rate Rd, a special end-capping agent E1 is introduced to adaptively terminate the incomplete chain link structure in G2 to generate a special polyether end product F with a functional group sequence order uniformity mean square deviation sigma of not higher than 10%; the functional group directional construction effect of the end product F is verified by using a two-dimensional nuclear magnetic resonance spectrum and a sequence recognition algorithm based on a machine learning training model, and the specific steps include: collecting the two-dimensional nuclear magnetic resonance spectrum of the end product F to obtain the coupling displacement signal between the functional groups and the cross-peak image; inputting the spectrum data into a trained sequence recognition model, wherein the sequence recognition model is constructed based on a combination of a convolutional neural network and a recurrent neural network, and a plurality of standard block arrangement samples are preset as a training set; outputting the probability distribution graph of the functional group position of each chain segment from the sequence recognition model, and comparing the similarity with a target construction template to determine the sequence construction deviation; when the similarity of the comparison result is higher than a 95% threshold, it is determined that the directional construction is successful, otherwise a structure reconstruction suggestion is triggered; the construction of the sequence recognition model comprises: performing image standardization processing on the two-dimensional nuclear magnetic resonance spectrum; inputting the processed spectrum into a spatial feature extraction module constructed based on a convolutional neural network to extract the local image features of the coupling relationship between the functional groups; the extraction result is further inputted into a sequence analysis module constructed based on a recurrent neural network to identify the functional group arrangement sequence in the block chain segment; in the sequence recognition model training process, a plurality of predefined polyether block arrangement maps are used as a supervised learning training set, the model weight is optimized through error back propagation, and finally the similarity matching result of the target product and the standard construction template is outputted; Based on the matrix, an average position shift Δp and a matching missing rate η are calculated, wherein Δp represents the mean difference between the actual position and the theoretical position of the functional group, and η represents the proportion of missed identification or mismatching; a construction deviation score is calculated by integrating Δp and η , which is expressed in a normalized score manner, with a full score of 100 points, and below 90 points is identified as significant construction deviation.

2. The method of directed construction of a specific functional special polyether material according to claim 1, characterized in that: determining the sequence construction deviation comprises: one-to-one corresponding comparison between the target polyether functional group arrangement sequence outputted by the sequence recognition model and the standard template sequence to construct a sequence matching mapping matrix; the molecular conformation parameters based on G1, which monitor the tension bending degree Delta theta between the molecular chain links in real time, comprise: scanning the real-time vibration peak of the block chain segment structure G1 in the reaction system by using an in-situ laser Raman spectrum to extract the characteristic frequency signal of the chain tension change; the characteristic frequency signal is converted into the tension bending degree Delta theta, and a tension-conformation correlation matrix is constructed through a Fourier transform algorithm. The Δθ parameter is input to the response directing domain P2 of M2, triggering the rotatable groups in P2 to undergo conformational adjustment, and re-directing the space of the insertion path of the functional group monomer B.

3. The method of directed construction of a specific functional special polyether material according to claim 2, characterized in that: The Δθ parameter is input to the response directing domain P2 of M2, triggering the rotatable groups in P2 to undergo conformational adjustment, including: The Δθ parameter is converted into a corresponding potential signal ε, and the degree of conformational deviation between the chain segments is reflected by the change in potential difference; The ε value acts on the response unit in P2, and the response unit includes a group of rotatable conformational groups connected by arylene-ether bonds, and the conformational angle of the group can be reversibly switched within ±45°; Under the driving of ε, the rotatable groups undergo conformational rearrangement, thereby changing the local polarity distribution of P2 and guiding the target functional group monomer B to approach the active recognition site on M1 along a new spatial path.

4. The method of claim 2, wherein the method is characterized by: The connection orientation of the B monomer on the main chain is automatically adjusted, including: During the polymerization reaction, the local torsional conformation and the chain segment strain direction of the block chain segment structure G1 are collected in real time, and the three-dimensional spatial orientation data of the molecular chain segment is obtained by a multi-channel angular distribution imager; Based on the spatial orientation data, the electrochromic response unit embedded in the auxiliary template M2 is activated, and the polarizable groups in the response directing domain P2 are driven to undergo reversible rotation or displacement, forming a guiding potential well; The B monomer is micro-captured by the potential well, so that it completes the connection in the conformational preferred direction when approaching the main chain, and ensures that the functional group is inserted into the main chain skeleton at the smallest conformational stress angle.

Images