Covalent bond functionalization optimization method for high-dielectric electric field sensitive heterojunction material

Abstract

The invention relates to the technical field of electric field sensitive materials, in particular to a covalent bond functionalization optimization method for a high-dielectric electric field sensitive heterojunction material. The method comprises the following steps: firstly, simulating and screening out a two-dimensional chalcogenide-graphene heterojunction substrate configuration with the highest dielectric polarization intensity and the optimal transmission performance by utilizing a first principle; performing oxidation functionalization treatment on the substrate according to the initial process parameters to introduce covalent bonding functional groups, and actually measuring Zeta potential and dielectric properties to calculate a comprehensive score; further constructing a Bayesian optimization model which takes the process parameters as input and takes the comprehensive score as a target, calculating an optimization direction by utilizing a Gaussian process and an expectation improvement function, and outputting a parameter correction value; and finally, carrying out closed-loop iteration of preparation, testing, evaluation and adjustment until the performance is converged. According to the invention, full-flow intelligent control from microstructure screening to macroscopic process optimization is realized, and the dielectric constant and dispersion stability of the heterojunction material are remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of electric field sensitive materials and sensors, and specifically relates to a method for optimizing the covalent bond functionalization of high dielectric electric field sensitive heterojunction materials. Background Technology

[0002] With the rapid development of the Internet of Things, flexible electronics, and industrial automation technologies, high-sensitivity electric field sensors are playing an increasingly crucial role in environmental monitoring, non-contact voltage detection, and smart devices. As the core medium of sensors, the dielectric properties and electron transport capabilities of electric field-sensitive materials directly determine the sensitivity and response speed of the devices. In existing technologies, two-dimensional transition metal chalcogenides, represented by molybdenum disulfide and tungsten disulfide, as well as graphene, have become the mainstream choice for constructing high-performance electric field-sensitive units due to their unique layered structure, tunable bandgap characteristics, and excellent carrier mobility. Researchers typically prepare these materials using processes such as chemical vapor deposition or mechanical exfoliation, and attempt to integrate the advantages of different materials by constructing heterojunctions to obtain composite substrates with better performance.

[0003] However, despite the enormous theoretical application potential of two-dimensional heterojunction materials, numerous technical bottlenecks remain in their practical preparation and functionalization. First, the intrinsic dielectric constant of these materials is generally limited, and graphene suffers from insufficient dielectric polarization due to the lack of polar groups. To address this, existing modification strategies often rely on physical mixing or non-covalent modification (such as introducing functional groups using van der Waals forces or π-π stacking). These non-chemically bonded interactions are weak, leading to extremely unstable adhesion of functionalized groups to the substrate surface. Desorption or uneven random distribution easily occurs, disrupting carrier migration pathways and potentially triggering severe phase separation. Furthermore, due to the surface energy mismatch at the heterojunction interfaces and the lack of strong chemical bonds, the materials readily aggregate in solvents, exhibiting poor dispersion stability (often resulting in low absolute Zeta potentials). This leads to numerous interfacial voids and defects in the shaped dielectric layer, severely impacting the long-term reliability of the sensor. More importantly, the current R&D model still mainly relies on the traditional "experience-based trial and error method," lacking theoretical guidance from the microscopic electronic structure level (such as band alignment and quantum capacitance), and the control of process parameters (such as temperature and time) lacks intelligent means, so even small process fluctuations can lead to huge differences in product performance.

[0004] Therefore, the urgent technical problem to be solved in the existing technology is how to overcome the defects of weak interfacial bonding, severe phase separation and poor dispersion caused by traditional physical mixing and non-covalent modification, and abandon the inefficient blind trial and error development mode. We need to establish a method that can accurately predict the microstructure of materials, establish stable chemical bonding interfaces, and systematically optimize complex process parameters, so as to significantly improve the dielectric constant and electric field response performance of heterojunction materials while ensuring the high uniformity and long-term stability of the materials. Summary of the Invention

[0005] The purpose of this invention is to address the above-mentioned shortcomings by providing a method for optimizing the covalent bond functionalization of high dielectric electric field sensitive heterojunction materials.

[0006] Firstly, a method for optimizing the covalent bond functionalization of high-dielectric-field-sensitive heterojunction materials is provided, employing the following technical solution: A method for optimizing covalent bond functionalization of high dielectric electric field sensitive heterojunction materials, the method comprising the following steps: Step (1): Obtain crystal structure data of graphene and group II chalcogenides, and use first-principles simulation to simulate the electronic structure characteristics of two-dimensional chalcogenide-graphene heterojunction; by simulating dielectric polarization intensity and carrier scattering rate under different stacking configurations, calculate and screen the heterojunction composition and stacking mode with the highest dielectric constant and the best electron transport performance, and output the best heterojunction substrate configuration. Step (2): Based on the optimal heterojunction substrate configuration output in step (1), prepare a two-dimensional chalcogenide-graphene heterojunction substrate; set the initial reaction temperature, reaction time, oxidant ratio and ultrasonic time as functionalization process parameters, and use the oxidation functionalization method to treat the heterojunction substrate to introduce oxygen-containing functional groups and obtain the functionalized heterojunction material to be tested. Step (3): Prepare the functionalized heterojunction material to be tested obtained in step (2) into a test sample, and actually collect Zeta potential data, as well as actual dielectric constant and loss tangent data at a preset frequency; based on the collected measured data, obtain the comprehensive score F through weighted function calculation, wherein the comprehensive score F is a weighted function of Zeta potential value and dielectric performance value; Step (4): Construct a Bayesian optimization model with reaction temperature, reaction time, oxidant ratio and ultrasonic time as input variables and comprehensive score F as objective function; use Gaussian process modeling, update the posterior distribution according to comprehensive score F, and obtain the optimization direction of objective function relative to input variables through expected improvement acquisition function calculation, and output the correction values ​​of reaction temperature, reaction time, oxidant ratio and ultrasonic time for the next iteration; Step (5): Based on the correction value output in step (4), adjust the functionalization process parameters in step (2) to re-prepare the functionalized heterojunction material, and repeat steps (2) to (4) until the comprehensive score F meets the convergence criterion, and output the final functionalization preparation process parameters.

[0007] Further, in step (1), the first principle adopts density functional theory, and the electronic structure characteristics include band structure, density of states and quantum capacitance; the two-dimensional chalcogenide-graphene heterojunction is selected from one of molybdenum disulfide / graphene, tungsten disulfide / graphene or molybdenum diselenide / graphene.

[0008] Furthermore, in step (2), the specific method for preparing the two-dimensional chalcogenide-graphene heterostructure substrate is chemical vapor deposition, which includes the following processes: In the tube furnace system, a mixture of argon and hydrogen is introduced, with the flow rate ratio controlled at 1~3:1 and the pressure maintained at 10Pa~100Pa. Using sulfur powder or selenium powder as a chalcogenide source and methane as a carbon source, the heterojunction substrate is deposited on the substrate at a temperature of 800℃~1000℃ by monitoring the partial pressure of the precursor.

[0009] Further, in step (2), the oxidative functionalization method includes the following process: dispersing the heterojunction substrate in a mixture of concentrated sulfuric acid and concentrated nitric acid, wherein the volume ratio of the mixture is 2.5~3.5:1; ultrasonically treating the dispersion under ice bath conditions of 0℃~5℃, adding potassium permanganate, transferring the reaction system to a constant temperature water bath, and continuously stirring the reaction; after the reaction is completed, centrifuging the product and washing it until neutral, and vacuum drying to obtain the functionalized heterojunction material to be tested.

[0010] Furthermore, in step (3), the actual data collection process includes: Zeta potential data were collected by observing the particle velocity through the application of an external electric field using a dynamic light scattering instrument, with the target absolute value set to be greater than 20mV. The dielectric constant was collected at a preset frequency of 1 kHz using an impedance analyzer with a parallel plate electrode configuration. and loss tangent ,in, target value , target value .

[0011] Further, in step (4), the construction and computation of the Bayesian optimization model includes: The Bayesian optimization process establishes a surrogate model through a Gaussian process, uses prior distribution and observed composite score F data to construct the probabilistic relationship between input variables and objective function, uses kernel function to capture nonlinear interactions between input variables, and periodically updates posterior distribution to reflect new experimental information; The expected improvement acquisition function is used for iterative optimization. The expected value of the objective function at the candidate point exceeds the current optimal value. This balances the search for unexplored areas and the utilization of known high-performance areas, thereby calculating the optimization direction and correction value.

[0012] Furthermore, in step (5), the convergence criterion is that the standard deviation of the comprehensive score F in multiple consecutive batches is controlled within 5%.

[0013] Furthermore, the final functionalized fabrication process parameters output after iterative optimization in step (5) converge to: The reaction temperature is 30℃~40℃; the reaction time is 2h~6h; the mass ratio of heterojunction substrate to oxidant is 1:3~7; and the ultrasonic time is 20min~40min.

[0014] Secondly, an electric field-sensitive heterojunction material employs the following technical solution: An electric field-sensitive heterojunction material, said material being prepared by the method described above; The material comprises a two-dimensional chalcogenide-graphene heterostructure substrate, and carboxyl and hydroxyl functional groups covalently bonded to the surface of the substrate; the dielectric constant of the material at a frequency of 1 kHz. Loss tangent Furthermore, the absolute value of the Zeta potential is greater than 20mV.

[0015] Thirdly, an electric field sensor employs the following technical solution: An electric field sensor includes electrodes and a dielectric layer; the dielectric layer is made of the aforementioned electric field-sensitive heterojunction material by spin coating or lamination process; the sensor is configured to detect weak electric field changes in the environment or industrial equipment.

[0016] The beneficial effects of this invention are: This invention provides a method for optimizing the covalent bond functionalization of high-dielectric-field-sensitive heterojunction materials. By deeply integrating first-principles calculations, covalent bond functionalization modification, and Bayesian optimization algorithms, the method first uses first-principles simulations to accurately predict the electronic structure and dielectric polarization characteristics of the heterojunction, selecting the optimal substrate configuration at the atomic scale. This theoretical guidance directly avoids the risk of band mismatch caused by blind trial and error, ensuring the material's high intrinsic dielectric potential. Based on this, stable carboxyl and hydroxyl functional groups are introduced onto the heterojunction surface using an oxidation functionalization method. By forming a robust covalent bond network, this effectively overcomes the defects of weak interfacial bonding and easy phase separation in traditional physical mixing or non-covalent modification, significantly improving the material's hydrophilicity and dispersion stability (high Zeta potential), thereby reducing carrier scattering and polarization losses caused by interfacial defects and voids. Furthermore, by constructing a Bayesian optimization model between process parameters and the overall material performance, and using a Gaussian process to capture the nonlinear relationship between parameters and perform closed-loop iterative optimization, this method can accurately lock the optimal window for reaction temperature, time, and oxidant ratio. This solves the problem that traditional single-factor experiments cannot simultaneously achieve multi-objective optimization. Thus, while ensuring high process repeatability, it maximizes the dielectric constant and minimizes the dielectric loss of the heterojunction material, significantly improving the overall performance and environmental adaptability of the electric field-sensitive material. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for optimizing the covalent bond functionalization of high-dielectric-field-sensitive heterojunction materials, as provided in an embodiment of the present invention.

[0018] Figure 2 The diagram shows the microscopic atomic model of a two-dimensional chalcogenide-graphene heterojunction constructed using first-principles calculations in this embodiment of the invention; wherein, (a) is a molybdenum diselenide / graphene heterojunction model, (b) is a molybdenum disulfide / graphene heterojunction model, and (c) is a tungsten disulfide / graphene heterojunction model.

[0019] Figure 3 The diagram shows the density of states (DOS) distribution of three different heterojunction systems calculated in the embodiments of the present invention; wherein, (a) corresponds to the molybdenum diselenide / graphene system, (b) corresponds to the molybdenum disulfide / graphene system, and (c) corresponds to the tungsten disulfide / graphene system. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the present application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on the present invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0021] In the following description, references to "some embodiments" refer to a subset of all possible embodiments; however, it is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. Unless otherwise defined, all technical and scientific terms used in the embodiments of the invention have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments of the invention pertain. The terminology used in the embodiments of the invention is for the purpose of describing the embodiments of the invention only and is not intended to limit the invention.

[0022] Those skilled in the art should understand that, in the following description of the embodiments of the present invention, the sequence of numbers does not imply the order of execution. Some or all steps may be executed in parallel or sequentially. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0023] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0024] Those skilled in the art will understand that the numerical ranges in the embodiments of the present invention should be understood to specifically disclose each intermediate value between the upper and lower limits of the range. Each smaller range between any stated value and an intermediate value within the stated range, as well as any other stated value or an intermediate value within the stated range, is also included within the present invention. The upper and lower limits of these smaller ranges may be independently included or excluded from the range.

[0025] Unless otherwise stated, the technical / scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. While only preferred methods and materials are described herein, any methods and materials similar or equivalent to those described herein may be used in embodiments or test cases of the invention. All references to this specification are generally incorporated herein by reference to disclose and describe methods and / or materials associated with said references. In the event of any conflict with any incorporated reference, the contents of this application shall prevail.

[0026] It should be noted that all raw materials and / or reagents in the embodiments of the present invention were purchased from the market or prepared according to conventional methods known to those skilled in the art.

[0027] This embodiment provides a method for optimizing the covalent bond functionalization of high-dielectric-field-sensitive heterojunction materials. This method constructs a closed-loop technology system encompassing the entire process, from first-principles calculations at the microscopic atomic scale, to chemical covalent bond functionalization modification at the mesoscopic scale, and finally to Bayesian intelligent iterative optimization of macroscopic process parameters. For example... Figure 1 As shown, this method solves the core technical challenges faced by traditional two-dimensional heterojunction materials in electric field sensing applications through systematic engineering design, such as low dielectric constant, weak interfacial bonding, poor dispersion stability, and reliance on trial and error in preparation processes. It achieves a synergistic improvement in the dielectric properties and electronic transport properties of heterojunction materials.

[0028] I. First-principles calculations and heterojunction substrate screening.

[0029] like Figure 1 The beginning part of the process and Figure 2 , Figure 3 As shown, this embodiment first performs step (1), which involves acquiring material microstructure data and performing simulation screening using first-principles calculations. This step aims to address the blind spots in material design from a theoretical perspective, and through precise quantum mechanical simulations, identify heterojunction configurations with optimal intrinsic dielectric polarization potential and electron transport characteristics.

[0030] In practice, initial crystal structure data of graphene and group II chalcogenides are first obtained from authoritative databases such as the Inorganic Crystal Structure Database (ICSD) or Materials Project. The preferred group II chalcogenides are molybdenum disulfide, tungsten disulfide, and molybdenum diselenide. Atomic models of two-dimensional chalcogenide-graphene heterostructures are then constructed using density functional theory-based computational software (such as VASP, Materials Studio's CASTEP, or the Dmol3 module).

[0031] During model construction, considering the special characteristics of two-dimensional materials, a sufficiently thick vacuum layer must be set in the normal direction (Z-axis) of the heterostructure plane. The thickness of this vacuum layer is typically set to 15 Å to 20 Å. Its technical significance lies in eliminating the non-physical electrostatic interactions between adjacent mirror layers caused by periodic boundary conditions, ensuring that the simulation results truly reflect the intrinsic properties of single-layer or few-layer heterostructures. Figure 2 As shown, molybdenum diselenide / graphene were constructed respectively. Figure 2 a) Molybdenum disulfide / graphene ( Figure 2 b) and tungsten disulfide / graphene ( Figure 2c) Microscopic heterojunction model. The ball-and-stick model in the figure clearly shows the arrangement of different atoms (e.g., gray represents carbon atoms, yellow / orange represents sulfur / selenium atoms, and cyan / blue represents molybdenum / tungsten atoms) at the interface. To reduce the interfacial stress caused by lattice mismatch, supercell matching technology is used, for example, constructing a 4×4 graphene unit cell and matching it with a 3×3 TMDs unit cell, and optimizing the lattice constant and atomic positions to achieve the lowest stable energy state of the system.

[0032] Subsequently, simulations were performed to calculate the electronic structure characteristics, with core parameters including band structure, density of states, and quantum capacitance. For example... Figure 3 As shown, the density of states distributions of the three heterojunction systems were calculated respectively. Figure 3 a corresponds to MoSe2 / G, Figure 3 b corresponds to MoS2 / G. Figure 3 c corresponds to WS2 / G). Fermi level in the density of states diagram ( The peak characteristics near the Fermi level are directly related to the material's electron transport capability and quantum capacitance. By integrating the density of states near the Fermi level, combined with the formula:

[0033] Quantitatively evaluate the charge storage capacity of different heterojunctions.

[0034] Furthermore, this embodiment focuses on simulating the dielectric polarization intensity under different stacking configurations. By changing the relative positions of the TMDs layers and graphene layers (e.g., AA stacking, AB stacking) or the interlayer torsion angle, the charge density difference under an applied electric field is calculated, thereby deriving the dielectric polarization tensor. Simultaneously, the carrier scattering rate is calculated based on deformation potential theory to evaluate the influence of interface phonon scattering on electron mobility. The system comprehensively compares the above calculation results and selects the heterojunction composition and stacking mode with the highest dielectric constant (strong polarization capability) and the lowest carrier scattering rate (low transmission loss), outputting it as the optimal heterojunction substrate configuration. For example, the calculation results may show that the MoS2 / G heterojunction exhibits the best overall performance under specific strain, and this will be used as the target substrate for subsequent experimental fabrication.

[0035] II. Preparation of two-dimensional heterojunction substrates.

[0036] Based on the optimal junction configuration (e.g., MoS2 / G) selected in step (1), the physical preparation stage of step (2) is then initiated. In this embodiment, a high-quality, large-area two-dimensional chalcogenide-graphene heterojunction substrate is prepared using chemical vapor deposition, a method that allows for precise control of the number of layers and crystal quality.

[0037] The specific implementation is carried out in a tube furnace equipped with a high-precision mass flow meter and a multi-zone temperature control system. First, high-purity copper foil or silicon dioxide / silicon (SiO2 / Si) wafers pre-grown with graphene are selected as the substrate. If copper foil is used, electrochemical polishing and high-temperature annealing are required to remove the surface oxide layer and increase the grain size.

[0038] The deposition process involves the following key process controls: sulfur powder (S) or selenium powder (Se) is placed as a chalcogen source in the upstream low-temperature zone of the tube furnace (typically 150℃~250℃); a metal precursor such as molybdenum trioxide (MoO3) or tungsten trioxide (WO3) is placed at a specific position before the downstream high-temperature zone; and the substrate is placed in the center of the high-temperature deposition zone. High-purity argon is introduced as the carrier gas, and a small amount of hydrogen is mixed in as a reducing agent and etchant. The argon to hydrogen flow ratio is controlled between 1:1 and 3:1 (e.g., Ar 50 sccm, H2 10 sccm) to maintain a reducing atmosphere and regulate the nucleation density.

[0039] The system pressure is maintained at a low level of 10 Pa to 100 Pa to increase the mean free path of gas molecules and promote uniform deposition. The heating program is set to raise the temperature in the high-temperature zone at a rate of 10°C / min to 20°C / min to 800°C to 1000°C (the specific temperature depends on the material system; for example, MoS2 is typically heated to 750°C to 850°C, while WS2 requires a higher temperature). Under this high-temperature environment, the chalcogenide source evaporates and is transported to the high-temperature zone by the gas flow. It reacts chemically with the reduced metal precursor vapor on the substrate surface and nucleates and grows laterally on the graphene lattice template through a van der Waals epitaxial growth mechanism, forming an atomically flat two-dimensional heterojunction film. By monitoring the partial pressure of the precursor (controlled by the source temperature) and the growth time in real time, the coverage and number of layers of the heterojunction are precisely controlled, ultimately obtaining the optimal heterojunction substrate with a complete structure and low defect density.

[0040] III. Oxidative functionalization of covalent bonds.

[0041] After obtaining a heterojunction substrate with a stable physical structure, in order to solve the problems of low surface energy, poor dispersibility in solvents and limited dielectric constant due to lack of polar groups, this embodiment performs the latter part of step (2), namely covalent functionalization modification.

[0042] In this stage, initial functionalization process parameters are set, including reaction temperature, reaction time, mass ratio of heterojunction to oxidant, and ultrasonic treatment time. This process employs a modified oxidative functionalization method (similar to a variant of the Hummers method, but with milder adjustments for two-dimensional heterojunctions), aiming to introduce oxygen-containing polar functional groups such as carboxyl groups (-COOH) and hydroxyl groups (-OH) into the carbon skeleton edges or chalcogenide defect sites on the heterojunction surface to form stable covalent bonds (CO, C=O, SO, etc.).

[0043] The specific operating procedure is as follows: First, a dispersion treatment is performed. The CVD-prepared heterojunction substrate is transferred from the substrate (e.g., via PMMA-assisted wet transfer) and washed, then dispersed in a strong acid medium. This medium is composed of concentrated sulfuric acid and concentrated nitric acid in a volume ratio of 2.5:1 to 3.5:1 (preferably 3:1). Concentrated sulfuric acid acts as a protonating agent, intercalating agent, and dehydrating agent, while concentrated nitric acid provides nitrocations as a strong oxidizing agent, capable of attacking the conjugated π-bond system.

[0044] Following this, ultrasonic activation is performed. To prevent the heat of dissolution generated when the strong acid mixes with the material from damaging the layered structure of the two-dimensional material, this step is strictly carried out under ice-water bath conditions of 0℃~5℃. The mixture is ultrasonically treated using a high-power ultrasonic instrument (such as a probe-type or bath-type ultrasonic instrument), and the duration is controlled by ultrasonic time parameters (e.g., initially set to 30 min). The cavitation effect of ultrasound can overcome the interlayer van der Waals forces, peeling the stacked heterojunctions into few-layer or single-layer nanosheets and exposing more reactive sites.

[0045] The oxidation reaction then proceeds. Potassium permanganate, acting as a strong oxidant, is slowly added while continuously stirring. The addition rate must be strictly controlled to prevent a violent reaction that could lead to temperature runaway. The mass ratio of the heterojunction substrate to potassium permanganate is determined by the oxidant ratio parameter (e.g., initially set at 1:5).

[0046] Potassium permanganate has an extremely high oxidation potential in acidic environments, enabling it to oxidize defects and deep edges on the surface of heterojunctions. After feeding, the reaction system is transferred from an ice bath to a constant-temperature water bath, and the temperature is raised to the set reaction temperature (e.g., 35°C). The reaction is then carried out with continuous stirring at this temperature, and the reaction time is controlled by the reaction time parameter (e.g., initially set to 4 hours). During this process, manganese ions are reduced, and a large number of oxygen-containing functional groups are grafted onto the surface of the heterojunction.

[0047] Finally, post-processing is performed. After the reaction is complete, the mixture is poured into a large amount of deionized water to terminate the reaction, and an appropriate amount of hydrogen peroxide is added to reduce the residual high-valence manganese ions. The solid product is separated by high-speed centrifugation (8000 rpm~10000 rpm) and repeatedly washed with dilute hydrochloric acid and deionized water until the pH of the supernatant is neutral. The precipitate is dried overnight in a vacuum drying oven (e.g., 60°C) to obtain the functionalized heterojunction material powder to be tested.

[0048] IV. Data Collection and Comprehensive Evaluation.

[0049] In order to objectively evaluate the effect of functional modification and provide real data feedback for subsequent algorithm optimization, this embodiment performs step (3), namely, actual data collection and evaluation.

[0050] First, the dried powder obtained in step (2) is redispersed in a solvent (such as N,N-dimethylformamide DMF or deionized water) to prepare a dilute dispersion. The Zeta potential is measured using a dynamic light scattering instrument combined with electrophoretic light scattering technology. The specific principle involves applying an external electric field, observing the electrophoretic mobility of charged particles in a liquid, and then calculating the Zeta potential using the Henry equation. The Zeta potential reflects the charge density and colloidal stability of the particle surface. In this invention, the goal is to obtain a Zeta potential with an absolute value greater than 20 mV (preferably greater than 30 mV), indicating that sufficient dissociative functional groups have been introduced into the surface, and the resulting electrostatic repulsion is sufficient to prevent nanosheet aggregation and ensure the uniformity of the material during subsequent film formation.

[0051] Secondly, the dielectric properties are evaluated. Functionalized heterojunction materials are composited with a polymer matrix (such as polyvinylidene fluoride PVDF or epoxy resin), or dense dielectric films are prepared through a filtration / pressing process. Using a precision impedance analyzer or LCR meter, along with a dielectric test fixture (parallel plate electrode configuration), the dielectric constant is measured at a preset frequency (preferably 1 kHz, as this is the typical operating frequency of electric field sensors). ) and loss tangent ( The dielectric constant reflects the polarization capability of a material under an electric field, mainly originating from the dipole polarization generated by the introduced polar functional groups and the Maxwell-Wagner interfacial polarization effect at the heterojunction interface. The loss tangent reflects the degree of energy loss in the conversion of electrical energy into heat energy. The optimization objective of this invention is to maximize... (Target value) Minimize at the same time (Target value) This allows for high-sensitivity and low-noise signal detection.

[0052] Based on the above measured data, a comprehensive score F is defined as a quantitative evaluation index. F is a weighted function, and its mathematical expression can be designed as follows:

[0053] in, The weighting coefficients are set according to the specific application's emphasis on dispersion, energy density, and losses. This comprehensive score F collapses the multi-dimensional performance indicators into a single scalar, which serves as the objective function value (Y value) for Bayesian optimization.

[0054] V. Bayesian Iterative Optimization.

[0055] To address the complex nonlinear relationship between multiple process parameters (temperature, time, ratio, ultrasound) and material properties (comprehensive score F), and the problem that the high experimental cost makes it unsuitable for large-scale full factorial experiments, this embodiment introduces steps (4) and (5), namely, Bayesian optimization based on artificial intelligence.

[0056] First, construct the optimization model. Define the input variable vector:

[0057] The target variable is the overall score. A Gaussian process is used as a surrogate model to fit the black-box relationship between the input variables and the objective function. The Gaussian process utilizes the prior distribution and observed data points (i.e., the process parameters from previous rounds of experiments and their corresponding measured F-values) to calculate the posterior distribution of the objective function in the unsampled region, i.e., the predicted mean and variance. For the kernel function, radial basis functions or Matern kernels are used to effectively capture the complex nonlinear interactions between parameters (e.g., high temperatures may require a shorter time frame, while low temperatures may require a longer time frame for coupling relationships).

[0058] Secondly, the optimization direction is calculated. The expected improvement acquisition function (EI) is used to guide the next experimental step. The EI function comprehensively considers "development" (i.e., finding better solutions in regions where current predicted values ​​are high) and "exploration" (i.e., finding potential solutions in regions with large prediction variance and high uncertainty). By maximizing the EI function, the algorithm outputs a new set of process parameter correction values ​​(e.g., suggesting adjusting the temperature from 35℃ to 38℃ and the oxidant ratio from 1:5 to 1:6).

[0059] Finally, closed-loop iteration and convergence are performed. The corrected values ​​output by the algorithm are fed back to step (2), and the preparation and functionalization experiments are repeated according to the new parameters. Then, the test in step (3) is performed to obtain a new comprehensive score F. The new data is then compared with... Add the dataset and update the Gaussian process model. Repeat the above "preparation-testing-modeling-recommendation" cycle. The convergence criterion is set as follows: the standard deviation of the comprehensive score F in multiple consecutive batches (e.g., 5 batches) is controlled within 5%, or the EI value is lower than the preset threshold, indicating that the global optimum has been found or the performance improvement space is minimal.

[0060] VI. Final Parameters and Product Characteristics.

[0061] After iterative optimization of the above system, the functionalization preparation process parameters finally locked in this embodiment converge to a specific range: reaction temperature is 30℃~40℃ (too low a temperature will result in insufficient oxidation, while too high a temperature will damage the heterojunction framework); reaction time is 2h~6h; the mass ratio of heterojunction substrate to oxidant is 1:3~7; and ultrasonic time is 20min~40min.

[0062] The high-dielectric-field-sensitive heterojunction material prepared using this optimal process exhibits the following microstructural characteristics: a high density of carboxyl (-COOH) and hydroxyl (-OH) functional groups are stably covalently bonded to the surface and edges of the two-dimensional chalcogenide-graphene heterojunction substrate. This structure endows the material with superior macroscopic properties: its dielectric constant at 1 kHz is [insert value here]. The dielectric loss is much higher than that of unmodified materials; It maintains good insulation properties; and the absolute value of the Zeta potential in the solvent is greater than 20 mV, exhibiting excellent dispersion stability and anti-settling ability.

[0063] VII. Sensor Applications.

[0064] This embodiment also includes applying the optimized material to an electric field sensor. The specific manufacturing process includes mixing a functionalized heterojunction material dispersion with a film-forming agent (such as PVA or PVDF), and then using spin coating, screen printing, or vacuum lamination to form a uniform and dense dielectric sensitive layer on the surface of interdigitated electrodes or parallel plate electrodes. Due to the material's high dielectric constant, the sensor exhibits extremely high polarization response sensitivity to changes in weak external electric fields. Simultaneously, the structural stability brought about by covalent functionalization makes the sensor less prone to phase separation or performance drift during long-term operation, making it particularly suitable for non-contact voltage monitoring of high-voltage transmission lines, industrial electrostatic detection, and environmental electromagnetic field sensing.

[0065] In summary, this specific implementation method achieves an efficient, controllable, and high-performance method for developing electric field-sensitive materials by combining "precise navigation" through first-principles calculations and "intelligent optimization" through Bayesian optimization with "powerful modification" through covalent bond functionalization.

[0066] Based on the aforementioned specific embodiments, in order to further verify the effectiveness of the "covalent bond functionalization method for high dielectric electric field sensitive heterojunction materials based on first principle calculation screening and Bayesian optimization" described in this invention, two sets of preferred embodiments (Example 1, Example 2) and four sets of comparative examples (Comparative Example 1, Comparative Example 2, Comparative Example 3, Comparative Example 4) are provided below.

[0067] General test conditions: All materials prepared in the examples and comparative examples were tested using the same standards. Zeta potential test: The material was dispersed in deionized water (concentration 0.1 mg / mL) and measured at 25°C using a Malvern Zetasizer Nano ZS instrument.

[0068] Dielectric property testing: The material was mixed with PVDF at a volume ratio of 40%, and hot-pressed into a disc with a diameter of 10 mm and a thickness of 0.5 mm. The dielectric constant was measured using a precision impedance analyzer at a frequency of 1 kHz. ) and loss tangent ( ).

[0069] Dispersion stability test: Let the dispersion stand for 72 hours and observe the sedimentation.

[0070] Example Example 1 This embodiment 1 uses the method of the present invention to prepare functionalized heterojunction materials through full-process optimization.

[0071] First-principles screening (step 1): Using VASP software for simulation, it was found that the interlayer coupling between molybdenum disulfide and graphene is the strongest and the theoretical quantum capacitance is the highest in the AB stack configuration. Therefore, MoS2 / G was selected as the substrate configuration.

[0072] Substrate preparation (step 2): High-quality MoS2 / G heterojunction substrates were grown at 850℃ using molybdenum trioxide and sulfur powder as precursors by CVD in an atmosphere of argon-hydrogen mixture (flow ratio 2:1).

[0073] Bayesian optimization and functionalization (steps 3-5): Initially, several sets of parameters are set for exploration, and input into the Bayesian optimization model. After 5 iterations, the model converges and outputs the optimal process parameters. Fabrication is then carried out according to these optimal parameters. The ratio of mixed acids is 3:1:concentrated sulfuric acid to concentrated nitric acid.

[0074] Ultrasound time: 30 minutes (under ice bath conditions).

[0075] Oxidizing agent ratio: heterojunction substrate: potassium permanganate = 1:5.

[0076] Reaction temperature: 35℃.

[0077] Reaction time: 4 hours. After the reaction was completed, the product was obtained by washing and drying.

[0078] Example 2 This embodiment 2 uses the method of the present invention to prepare functionalized heterojunction materials through full-process optimization.

[0079] First-principles screening (step 1): Calculations show that WS2 / G has a lower carrier scattering rate in specific high-frequency applications, and it is selected as the substrate.

[0080] Substrate preparation (step 2): The CVD process temperature was adjusted to 920℃, and the rest was the same as in Example 1.

[0081] Bayesian optimization and functionalization (steps 3-5): Since WS2 has slightly higher chemical stability than MoS2, the Bayesian optimization algorithm automatically adjusts the parameters and eventually converges to: Oxidizing agent ratio: heterojunction substrate: potassium permanganate is 1:6.

[0082] Reaction temperature: 38℃.

[0083] Reaction time: 5 hours. The remaining steps are the same as in Example 1.

[0084] Comparative Example Comparative Example 1 Comparative Example 1 aims to demonstrate the original material properties without the introduction of covalent functional groups.

[0085] Substrate preparation: The MoS2 / G heterojunction was prepared in exactly the same manner as in Example 1.

[0086] Post-treatment: No acid treatment or potassium permanganate oxidation is performed; only simple solvent ultrasonic dispersion is carried out.

[0087] Comparative Example 2 Comparative Example 2 employs a physical modification method commonly used in the prior art.

[0088] Materials: The MoS2 / G heterojunction from Example 1 was used.

[0089] Modification method: No oxidizing agent is used. The material is dispersed in an aqueous solution containing sodium dodecylbenzenesulfonate (SDBS, a surfactant), and ultrasonically treated for 1 hour to achieve physical adsorption modification using van der Waals forces.

[0090] Comparative Example 3 Comparative Example 3 did not undergo Bayesian optimization and blindly adopted strong oxidation conditions.

[0091] Substrate: Same as in Example 1.

[0092] Functionalization parameters: reaction temperature: 60℃; oxidant ratio: 1:10; reaction time: 12h.

[0093] Comparative Example 4 Comparative Example 4 was not optimized using Bayesian methods and had mild oxidation conditions.

[0094] Substrate: Same as in Example 1.

[0095] Functionalization parameters: reaction temperature: 10℃; oxidant ratio: 1:1; reaction time: 1h.

[0096] III. Experimental Results and Analysis The materials prepared in the above embodiments and comparative examples were subjected to performance tests, and the results are shown in Table 1 below.

[0097] Table 1. Performance data of the materials provided in Examples 1-2 and Comparative Examples 1-4

[0098] As shown in Table 1, the dielectric constant of Example 1 (58.4) is much higher than that of Comparative Example 1 (12.5). This is because the introduced carboxyl and hydroxyl groups are polar groups, which provide a significant dipole polarization contribution under an electric field; at the same time, the Zeta potential increased from -12.4 mV to -38.5 mV, indicating that covalent bonding significantly improved the dispersibility.

[0099] Compared to Comparative Example 2, although physical mixing can also improve the dielectric constant to some extent (22.8), its dielectric loss is higher (0.085 vs 0.038). This is because the interfacial bonding force of physical adsorption is weak, which easily generates interfacial voids and defects, leading to energy dissipation. In contrast, the chemical interface formed by covalent bonds in Example 1 is more compact, reducing the loss of Maxwell-Wagner interfacial polarization.

[0100] Although Comparative Example 3 has a high dielectric constant (62.3), its loss tangent deteriorates sharply (0.450), and TEM observation shows that the crystal structure is destroyed. This is because the excessive oxidation conditions break the conjugated framework of molybdenum disulfide and graphene, resulting in an abnormal increase in the material's conductivity (increased leakage current), which makes it lose its value as an excellent dielectric layer (overall score F is only 0.40).

[0101] The performance improvement of Comparative Example 4 is limited. =18.6), and the low Zeta potential (-15.8 mV) indicates insufficient functional group grafting density, failing to achieve the modification objective. Example 1 is right at the performance sweet spot: the Bayesian optimization algorithm locked parameters such as 35℃ and a 1:5 ratio, which introduced a sufficient amount of functional groups (high... It achieves high Zeta content while preserving the integrity of the heterojunction framework (low loss), thus optimizing overall performance.

[0102] Example 2 shows that the method is also effective for the WS2 / G system. By automatically fine-tuning the process parameters using a Bayesian algorithm, materials with high dielectric constant (52.1) and low loss (0.032) that meet the requirements were also obtained.

[0103] Experimental data fully demonstrate that the method proposed in this invention can prepare heterojunction materials that possess high dielectric constant, low dielectric loss, and excellent dispersion stability. In particular, the Bayesian optimization step effectively avoids structural damage caused by excessive oxidation and mediocre performance caused by insufficient oxidation, ensuring that the process parameters fall within the optimal window, demonstrating significant technological advancement.

[0104] For those skilled in the art, other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations, but obvious variations or modifications derived therefrom are still within the scope of protection of the claims of this invention.

Claims

1. A method for optimizing the covalent bond functionalization of high-dielectric-field-sensitive heterojunction materials, characterized in that, The method includes the following steps: Step (1): Obtain crystal structure data of graphene and group II chalcogenides, and use first-principles simulation to simulate the electronic structure characteristics of two-dimensional chalcogenide-graphene heterojunction; by simulating dielectric polarization intensity and carrier scattering rate under different stacking configurations, calculate and screen the heterojunction composition and stacking mode with the highest dielectric constant and the best electron transport performance, and output the best heterojunction substrate configuration. Step (2): Based on the optimal heterojunction substrate configuration output in step (1), prepare a two-dimensional chalcogenide-graphene heterojunction substrate; set the initial reaction temperature, reaction time, oxidant ratio and ultrasonic time as functionalization process parameters, and use the oxidation functionalization method to treat the heterojunction substrate to introduce oxygen-containing functional groups and obtain the functionalized heterojunction material to be tested. Step (3): Prepare the functionalized heterojunction material to be tested obtained in step (2) into a test sample, and actually collect Zeta potential data, as well as actual dielectric constant and loss tangent data at a preset frequency; based on the collected measured data, obtain the comprehensive score F through weighted function calculation, wherein the comprehensive score F is a weighted function of Zeta potential value and dielectric performance value; Step (4): Construct a Bayesian optimization model with reaction temperature, reaction time, oxidant ratio and ultrasonic time as input variables and comprehensive score F as objective function; use Gaussian process modeling, update the posterior distribution according to comprehensive score F, and obtain the optimization direction of objective function relative to input variables through expected improvement acquisition function calculation, and output the correction values ​​of reaction temperature, reaction time, oxidant ratio and ultrasonic time for the next iteration; Step (5): Based on the correction value output in step (4), adjust the functionalization process parameters in step (2) to re-prepare the functionalized heterojunction material, and repeat steps (2) to (4) until the comprehensive score F meets the convergence criterion, and output the final functionalization preparation process parameters.

2. The method according to claim 1, characterized in that, In step (1), the first principle adopts density functional theory, and the electronic structure characteristics include band structure, density of states and quantum capacitance; the two-dimensional chalcogenide-graphene heterojunction is selected from one of molybdenum disulfide / graphene, tungsten disulfide / graphene or molybdenum diselenide / graphene.

3. The method according to claim 1, characterized in that, In step (2), the specific method for preparing the two-dimensional chalcogenide-graphene heterostructure substrate is chemical vapor deposition. Includes the following processes: In the tube furnace system, a mixture of argon and hydrogen is introduced, with the flow rate ratio controlled at 1~3:1 and the pressure maintained at 10Pa~100Pa. Using sulfur powder or selenium powder as a chalcogenide source and methane as a carbon source, the heterojunction substrate is deposited on the substrate at a temperature of 800℃~1000℃ by monitoring the partial pressure of the precursor.

4. The method according to claim 1, characterized in that, In step (2), the oxidative functionalization method includes the following process: dispersing the heterojunction substrate in a mixture of concentrated sulfuric acid and concentrated nitric acid, wherein the volume ratio of the mixture is 2.5~3.5:1; ultrasonically treating the dispersion under ice bath conditions of 0℃~5℃, adding potassium permanganate, transferring the reaction system to a constant temperature water bath, and continuously stirring the reaction; after the reaction is completed, centrifuging the product and washing it until neutral, and vacuum drying to obtain the functionalized heterojunction material to be tested.

5. The method according to claim 1, characterized in that, In step (3), the actual data collection process includes: Zeta potential data were collected by observing the particle velocity through the application of an external electric field using a dynamic light scattering instrument, with the target absolute value set to be greater than 20mV. The dielectric constant was collected at a preset frequency of 1 kHz using an impedance analyzer with a parallel plate electrode configuration. and loss tangent ,in, target value , target value .

6. The method according to claim 1, characterized in that, In step (4), the construction and computation of the Bayesian optimization model includes: The Bayesian optimization process establishes a surrogate model through a Gaussian process, uses prior distribution and observed composite score F data to construct the probabilistic relationship between input variables and objective function, uses kernel function to capture nonlinear interactions between input variables, and periodically updates posterior distribution to reflect new experimental information; The expected improvement acquisition function is used for iterative optimization. The expected value of the objective function at the candidate point exceeds the current optimal value. This balances the search for unexplored areas and the utilization of known high-performance areas, thereby calculating the optimization direction and correction value.

7. The method according to claim 1, characterized in that, In step (5), the convergence criterion is that the standard deviation of the comprehensive score F in multiple consecutive batches is controlled within 5%.

8. The method according to claim 1, characterized in that, The final functionalized fabrication process parameters output after iterative optimization in step (5) converge to: The reaction temperature is 30℃~40℃; the reaction time is 2h~6h; the mass ratio of heterojunction substrate to oxidant is 1:3~7; and the ultrasonic time is 20min~40min.

9. An electric field-sensitive heterojunction material, characterized in that, The material is prepared by the method according to any one of claims 1 to 8; The material comprises a two-dimensional chalcogenide-graphene heterostructure substrate, and carboxyl and hydroxyl functional groups covalently bonded to the surface of the substrate; the dielectric constant of the material at a frequency of 1 kHz. Loss tangent Furthermore, the absolute value of the Zeta potential is greater than 20mV.

10. An electric field sensor, characterized in that, It includes electrodes and a dielectric layer; the dielectric layer is made of the electric field-sensitive heterojunction material as described in claim 9, and is prepared by spin coating or lamination process; the sensor is configured to detect weak electric field changes in the environment or industrial equipment.

Images