A robot-assisted digitally controllable method for synthesizing nanocrystal morphology

Abstract

The present invention relates to a method for robot-assisted digital controllable synthesis of nanocrystal morphology, and specifically discloses a robot-assisted digital controllable synthesis of nanocrystal morphology modeling method, which includes the following steps: S1) using a high-throughput experimental method to obtain experimental conditions for preparing nanocrystals and the longitudinal plasma resonance absorption peak (LSPR) of the prepared nanocrystals to form a database; S2) screening the experimental data in the training database, determining an experimental condition as a variable, and forming a training database of the variable and the corresponding LSPR; S3) using a machine learning algorithm to obtain a thermodynamic model for digital manufacturing of nanocrystals; S4) determining the experimental conditions for preparing target nanocrystals based on the obtained fitting curve and the LSPR of the target nanocrystals; wherein the experimental data of the nanocrystals include the type and amount of the prepared raw materials. The present invention realizes the rational digital synthesis of nanocrystal materials through robot assistance.

Description

Technical Field

[0001] The present invention belongs to the field of digital intelligent manufacturing of nanocrystal materials, and specifically relates to a thermodynamic model and a modeling method for nanocrystal synthesis. Background Art

[0002] With the development of material genome and artificial intelligence technology, data-driven scientific discovery is becoming the "fourth research paradigm" after the "experimental paradigm", "theoretical paradigm" and "simulation paradigm". The coordinated use of digital technology, digital thinking and digital cognition to explore the cross-integration of the fourth paradigm and materials science will provide a new methodology for the development of new material preparation technology and digital manufacturing. Recently, top international journals such as Nature and Science have published articles that believe that programmable material preparation [1] research is a major scientific breakthrough produced through cross-disciplinary research.

[0003] However, digital intelligent automatic material preparation technology has only been applied to the development of living biological materials[2], organic chemical materials[1,3,4], small molecule drugs[5], polymers[6,7] and other materials. There has been no report on the related work of robot-assisted programmable preparation of nanocrystals.

[0004] Another key factor hindering the digitization of materials is the lack of a universal programming language. et al. [8] developed the first molecular programming language CRN++ suitable for synthetic biology. They used the working principle of computers, cells as hardware and genes as software to assemble new artificial biomaterials. Subsequently, Lu, Ellis et al. [2] reported the growth of programmable living biomaterials from engineered microbial co-cultures. In terms of organic chemical materials, Cronin et al. successively reported a programming language-driven organic material synthesis robot system (named Chemputer) [1], a standard operating system that can autonomously learn literature and automatically perform organic chemical synthesis [3], and a universal programmable chemical synthesis machine that can perform different organic syntheses [4]. Segler, Waller et al. reported the use of coded artificial intelligence to achieve retrosynthesis of organic small molecule drugs [5]. Zhu, Xu, Tan et al. reported a programmable polymer library and its logic gates that can perform biological logic operations. Gate)[6]; Xie Tao's team discovered the light-triggered topological programmability of dynamic covalent polymer networks[7]. For metal organic framework materials (MOF), Yaghi, Li et al. revealed that programmable chemical synthesis can be achieved by adjusting the sequence of metals such as cobalt, cadmium, lead and manganese in MOF-74[9]. In addition, Yu Shuhong's team developed a programmable preparation of graphene-based composite aerogel materials based on the melamine-sponge-templated hydrothermal process

[10] .

[0005] However, there are few reports on robot-assisted digital controllable synthesis of nanocrystal morphology and models for studying the thermodynamic mechanism of nanocrystal growth process through high-throughput big data.

[0006] References

[0007] 1.Steiner,S.,et al.,Organic synthesis in a modular robotic systemdriven by a chemical programming language.Science,2019.363(6423):p.eaav2211.

[0008] 2.Gilbert,C.,et al.,Living materials with programmablefunctionalities grown from engineered microbial co-cultures.Nature Materials,2021.

[0009] 3.Mehr,S.H.M.,et al.,A universal system for digitization andautomatic execution of the chemical synthesis literature.Science,2020.370(6512):p.101-108.

[0010] 4.Angelone,D.,et al.,Convergence of multiple synthetic paradigms in auniversally programmable chemical synthesis machine.Nature Chemistry,2021.13(1):p.63-69.

[0011] 5.Segler,M.H.S.,M.Preuss,and M.P.Waller,Planning chemical syntheseswith deep neural networks and symbolic AI.Nature,2018.555(7698):p.604-610.

[0012] 6.Zhang,P.,et al.,A programmable polymer library that enables theconstruction of stimuli-responsive nanocarriers containing logic gates.NatureChemistry,2020.12(4):p.381-390.

[0013] 7.Zou,W.,et al.,Light-triggered topological programmability in adynamic covalent polymer network.Science Advances,2020.6(13):p.eaaz2362.

[0014] 8. ,M.,Soloveichik,D.&Khurshid,S.,CRN++:Molecular programming language.Natural Computing,2020.19:p.391–407.

[0015] 9. Ji, Z., T. Li, and OMYaghi, Sequencing of metals in multivariate metal-organic frameworks. Science, 2020.369(6504): p.674-680.

[0016] 10.Ge,J.,et al.,A General and Programmable Synthesis of Graphene-Based Composite Aerogels by a Melamine-Sponge-Templated HydrothermalProcess.CCS Chemistry,2020.2(2):p.1-12. Summary of the invention

[0017] In order to overcome the problems of the prior art, the present invention, based on high-throughput experiments, constructs a thermodynamic model of nanocrystals through digital manufacturing of nanocrystalline materials and machine learning, and realizes controllable synthesis of nanocrystals based on the obtained thermodynamic model.

[0018] Therefore, in order to further study the Wulff and inverse Wulff thermodynamic models and their programmable languages, especially the rational preparation of colloidal gold nanorods, software technical support such as models, databases and algorithms is provided, which is the key to realizing programmable nanocrystal digital manufacturing.

[0019] The purpose of the present invention is to provide a thermodynamic model for nanocrystal preparation and its algorithm and programmable language, which solve the key common scientific problems existing in the prior art such as rational design, preparation and characterization preparation of new materials and the problem of lack of digital programming language.

[0020] One aspect of the present invention provides a method for modeling a thermodynamic model of digitally manufactured nanocrystals, comprising the following steps:

[0021] S11) using a high-throughput experimental method to obtain experimental conditions for preparing nanocrystals and LSPRs of the prepared nanocrystals to form a database;

[0022] S12) screening the experimental data in the training database, determining one experimental condition as a variable, and fixing other experimental conditions to form a training database of variables and corresponding LSPRs;

[0023] S13) using a machine learning algorithm SISSO to fit the fitting curve between the experimental conditions of the nanocrystals determined in step S12) and the LSPR of the nanocrystals to obtain a thermodynamic model for digital manufacturing of the nanocrystals;

[0024] S14) determining the experimental conditions for preparing the target nanocrystals according to the obtained fitting curve and the LSPR of the target nanocrystals;

[0025] Among them, the experimental data of nanocrystals include the types and amounts of raw materials used in preparation.

[0026] Another aspect of the present invention provides a prediction method for digital manufacturing of nanocrystals, comprising: obtaining the LSPR value of the nanocrystals to be prepared, and obtaining corresponding reaction conditions through the thermodynamic model of digital manufacturing of nanocrystals of the present invention.

[0027] Another aspect of the present invention provides a method for constructing a model of the quantitative relationship between the ratio of the surface energy of the crystal face of a nanocrystal and the silver ion concentration of the reaction system, thereby obtaining the thermodynamic relationship between the silver ion concentration and the ratio of the crystal face surface energy or the crystal aspect ratio, which comprises the following steps:

[0028] S21) screening the crystal morphology and crystal plane data of the target crystal in the crystallographic database constructed by Wulff;

[0029] S22) analyzing whether there is a trend between the surface area of ​​different crystal planes and the aspect ratio of nanorods;

[0030] S23) Select a surface energy ratio that has a trend with the aspect ratio of the nanorods (such as the ratio of the surface energies γ(110) and γ(001) of the (110) crystal plane to the (001) crystal plane), and construct a classical model and a machine learning model (using LSPR to reflect the aspect ratio value of the crystal morphology) between the surface energy ratio of the crystal plane and the silver ion concentration of the reaction system.

[0031] Another aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the steps of the modeling method of the thermodynamic model for modeling the digital manufacturing of nanocrystals described in the present invention.

[0032] In another aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the program, the steps of the modeling method of the thermodynamic model of digital manufacturing of nanocrystals are implemented.

[0033] Beneficial Effects

[0034] 1) This invention provides the first example of the fourth paradigm of data-driven scientific discovery and materials science, especially the cross-integration with the preparation of nanocrystalline materials, providing a new method for the development of new concept materials and common materials science.

[0035] 2) The present invention deeply studies the thermodynamic model of nanocrystals, provides models, databases, algorithms and programmable languages, and realizes programmable digital manufacturing of nanocrystals. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 . Schematic diagram of the process of using machine learning to construct a thermodynamic model of gold nanorod growth.

[0037] Figure 2 .Various crystal equilibrium morphologies of gold nanocrystals.

[0038] Figure 3 .Correlation analysis between the surface area and aspect ratio of the five equivalent surfaces of the (A)(100) surface of gold nanocrystals. (A)(100) crystal surface, A 100 =A (100) +A (I00) +A (010) +A (0I0) , (B) (110) crystal plane, A 110 =A (110) +A (I10) +A (110) +A (1I0) , (C) (111) crystal plane, A 111 =A (111) +A(I11) +A (1I1) +A (11I) +A (II1) +A (1II) +A (I1I) +A (III) , (D)(011) crystal plane, A 011 =A (011) +A (101) +A ( 0 I1) +A (01I) +A (I01) +A (10I) +A (I0I) +A (0II) , (E)(001) crystal plane, A 001 =A (001) +A (00I) .

[0039] Figure 4 . Transmission electron microscope images and size distribution of gold nanorods. (AC) The corresponding LSPR peaks of gold nanorods are 630nm, 784nm and 812nm. (DF) The average diameter and length of gold nanorods. The conversion formula of LSPR and aspect ratio (AR) is: AR = (LSPR-418) / 96

[0040] Figure 5 .Fitting curve of the classic model for nanocrystal synthesis.

[0041] Figure 6 . Machine learning curve fitting for nanocrystal synthesis. DETAILED DESCRIPTION

[0042] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and understandable, the specific implementation modes of the present invention are described in detail below, but it should not be understood as limiting the applicable scope of the present invention.

[0043] A specific embodiment of the present invention provides a method for modeling a thermodynamic model of digital manufacturing of nanocrystals, which includes the following steps:

[0044] S11) using a high-throughput experimental method to obtain experimental conditions for preparing nanocrystals and LSPRs of the prepared nanocrystals to form a training database;

[0045] S12) screening the experimental data in the training database, determining one experimental condition as a variable, and fixing other experimental conditions;

[0046] S13) using a machine learning algorithm SISSO to fit the fitting curve between the experimental conditions of the nanocrystals determined in step S12) and the LSPR of the nanocrystals to obtain a thermodynamic model for digital manufacturing of the nanocrystals;

[0047] S14) determining the experimental conditions for preparing the target nanocrystals according to the obtained fitting curve and the LSPR of the target nanocrystals;

[0048] Among them, the experimental data of nanocrystals include the types and amounts of raw materials used in preparation.

[0049] In a specific embodiment of the present invention, S12) comprises:

[0050] S121) screening the experimental data in the database, selecting one of the experimental conditions as a variable, and setting the other experimental conditions to fixed values;

[0051] S122) fitting the nanocrystal classical model based on the variables screened in S121) and the corresponding LSPR values ​​of the nanocrystals;

[0052] S123) Repeat steps S121) and S122) to obtain the degree of fit between different variables and the classical model of nanocrystals, and select the variable with the highest degree of fit to generate a training database.

[0053] In step S122), the nanocrystal classical model is a parameter equation obtained by fitting based on the Gibbs adsorption isotherm and the Langmuir adsorption isotherm. The Gibbs adsorption isotherm is: The Langmuir adsorption isotherm is: The parameter equation obtained by fitting is γ=e 0 In(1+ce 1 )+e 2 .

[0054] In the above specific embodiments, the nanocrystals are selected from gold nanocrystals.

[0055] In a specific embodiment of the present invention, the nanocrystals are gold nanocrystals. The method for preparing the gold nanocrystals is to use HAuCl 4 , CTAB, AgNO 3 , ascorbic acid, hydrochloric acid and sodium borohydride to react to obtain rod-shaped gold nanocrystals. The gold salt is selected as chloroauric acid solution. The surfactant is selected as CTAB. By screening different reaction raw materials and fitting with the classical model, a total of 6 raw materials were screened, AgNO 3The concentration of the solution, when used as the reaction raw material, the experimental results have the highest degree of fit with the classical model, and AgNO is selected 3 The concentration of the solution was used as a variable in the training set;

[0056] The thermodynamic model curve expression is LSPR = (3.625 × ln (C (Ag + ))^ 2 -3.43×(C(Ag + ))+0.58×(C(Ag + ))×ln(C(Ag + ))+6.31)×96+418, where the parameter corresponding to the ratio of the crystal surface energy is aspect ratio AR=3.625×ln(C(Ag + ))^ 2 -3.43×(C(Ag + ))+0.58×(C(Ag + ))×ln(C(Ag + ))+6.31; C(Ag + ) represents the silver ion concentration.

[0057] Another specific embodiment of the present invention provides a method for constructing a model of the thermodynamic relationship between the surface area of ​​a nanocrystal face and the surface energy of the crystal face, which comprises the following steps:

[0058] S21) screening the crystal morphology and crystal surface data of the target crystal in the crystallographic database constructed by Wulff;

[0059] S22) analyzing whether there is a trend between the surface area of ​​different crystal planes and the aspect ratio of nanorods;

[0060] S23) selecting a surface energy ratio that has a trend with the aspect ratio of the nanorods, and constructing a classical model and a machine learning model of the crystal surface energy ratio and the reaction system and reaction conditions.

[0061] Specifically, the target crystal is a gold nanocrystal, and the crystallographic database constructed by Wulff is selected from a cubic crystal system database.

[0062] Specifically, the surface energy ratio in step S23) is, for example, the ratio of the surface energies γ(110) and γ(001) of the (110) crystal plane to the (001) crystal plane.

[0063] Specifically, the reaction conditions, for example, the concentration of silver ions is used as the reaction condition in the preparation process of gold nanocrystals.

[0064] Specifically, the crystal aspect ratio is expressed in the form of LSPR numerical value, and the conversion formula between LSPR and aspect ratio (AR) is: AR=(LSPR-418) / 96.

[0065] In step S23), an artificial neural network model between the crystal surface area and the crystal surface energy and the crystal surface energy ratio is obtained by using the crystal surface energy ratio as a descriptor and the surface energy ratio as an output item through an artificial neural network machine learning method.

[0066] The present invention is implemented by the following method:

[0067] Example 1 Establishment of a thermodynamic relationship model between gold nanosurface morphology and surface energy

[0068] Taking gold nanoparticles as an example, the first step is to call up seven crystal system databases in crystal structure, namely cubic, hexagonal, triangular, tetragonal, orthorhombic, monoclinic, and triclinic. The database is based on the Wulff structure and contains more than 2,000 possible crystal morphologies and 90,000 different crystal face data. In the cubic crystal system data, a variety of possible gold nanocrystal morphologies were screened. According to the Wulff theorem, the surface energy of the (hkl) surface is proportional to the distance from the center of the crystal to the corresponding surface:

[0069] d hkl ~γ hkl

[0070] The various crystal equilibrium morphologies of gold nanocrystals obtained by Wulff construction are as follows Figure 2 In order to rationally design the target nanocrystals, the simulated morphology was quantitatively analyzed using typical geometric features such as specific surface area and aspect ratio. According to the symmetry of the cubic crystal system, five equivalent crystal planes were identified, and the relevant analysis is shown in Figure 3 As shown. Figure 3 E, it is found that the surface area of ​​the (001) crystal plane gradually decreases with the increase of the aspect ratio of the nanorods. From the results of transmission electron microscopy (TEM), the length, diameter and surface area information of the gold nanorods were obtained using the image reading code. Figure 2 and Figure 4 The TEM morphology of Au u nanorods (AR = 2.2, LSPR = 630 nm) is shown. However, the surface area can easily change due to some unpredictable morphological conditions. The biggest challenge for experimental studies is to clarify the multiple surface regions corresponding to different potential morphologies.

[0071] The present invention uses the surface area of ​​the (110) or (001) crystal plane and the surface energy γ(001) / γ(110) ratio of the crystal plane as descriptors, and uses the surface energy ratio of the above crystal plane as the output target value, and uses an artificial neural network (ANN) model to accelerate the calculation of the surface energy ratio. The experimental results show that the predicted surface energy ratio of machine learning is in good agreement with the calculated value of Wulff construction. 2=0.99. Therefore, the present invention uses a machine learning method to verify the thermodynamic relationship model between the silver ion concentration and the surface energy of the gold nanorod synthesis system: Y = e 0 In(1+ce 1 )+e 2 .

[0072] Example 2 Thermodynamic Model of Controllable Synthesis Driven by Thermodynamic Data

[0073] Using the classical mathematical model, we first studied the correlation between the calculated surface energy of the prepared gold nanocrystals and the preliminary experimental results. and Langmuir adsorption isotherm The parametric equation Y = e was established 0 In(1+ce 1 )+e 2 (classical model), such as Figure 5 As shown in the figure, the surface energy detection range of the fitting curve in the classical model corresponds to the LSPR value range of 666-878nm, and the surface energy variance between the fitting curve and the actual experiment is 0.98. Furthermore, by taking advantage of the computer-assisted high-throughput experimental equipment of the present invention, a large data set containing LSPR information and experimental methods was obtained. The morphology regulator silver ion in the reaction raw material, i.e., AgNO 3 The concentration of AgNO was taken as a variable and the machine learning algorithm SISSO was used to fit the AgNO 3 The fitting curve between the concentration and LSPR is shown in Figure 6 Through the machine learning model, not only the AgNO 3 The application range of the factor is expanded, and the LSPR value range is extended to 600-925nm; and the variance after fitting with the experimental value is more accurate than the classical model.

[0074] The present invention applies the reverse Wulff structure to the colloidal nanogold system and determines the surface energy and morphology regulation and Ag + In order to verify the applicability of this method, the inverse Wulff structure was applied to the equilibrium morphology of gold nanorods observed experimentally to infer the surface energy, which was used to determine the growth conditions of a given gold nanorod crystal morphology. In the present invention, the surface energy is related to the important solution parameter Ag. + concentration, considering different Ag +The influence of concentration on the morphology of gold nanorods. The existing experimental results are extrapolated to unknown experimental conditions, which makes it possible to predict the morphology of gold nanorods under certain experimental conditions. The present invention uses machine learning to construct a gold nanorod growth model, which has a high degree of consistency and a wide prediction range compared with the inverse Wulff construction calculation model. The surface energy obtained from the experimentally observed particle morphology is obtained by minimization algorithm, and the obtained surface energy is consistent with the actual growth conditions.

Claims

1. A method for modeling a thermodynamic model for digital manufacturing of nanocrystals. It is characterized in that It includes the following steps: S11) using a high-throughput experimental method to obtain experimental conditions for preparing nanocrystals and LSPRs of the prepared nanocrystals to form a database; S12) screening the experimental data in the training database, determining one experimental condition as a variable, and fixing other experimental conditions to form a training database of variables and corresponding LSPRs; S13) using a machine learning algorithm SISSO to fit the fitting curve between the experimental conditions of the nanocrystals determined in step S12) and the LSPR of the nanocrystals to obtain a thermodynamic model for digital manufacturing of the nanocrystals; S14) determining the experimental conditions for preparing the target nanocrystals according to the obtained fitting curve and the LSPR of the target nanocrystals; Among them, the experimental data of nanocrystals include the types and amounts of raw materials used in preparation.

2. The modeling method according to claim 1, It is characterized in that Step S12) comprises: S121) screening the experimental data in the database, selecting one of the experimental conditions as a variable, and setting the other experimental conditions to fixed values; S122) fitting the nanocrystal classical model based on the variables screened in S121) and the corresponding LSPR values ​​of the nanocrystals; S123) Repeat steps S121) and S122) to obtain the degree of fit between different variables and the classical model of nanocrystals, and select the variable with the highest degree of fit to generate a training database.

3. The modeling method according to any one of claims 1 to 2, It is characterized in that The nanocrystals are selected from gold nanocrystals.

4. The modeling method according to any one of claims 1 to 2, It is characterized in that The nanocrystals are gold nanocrystals; The method for preparing gold nanocrystals is to use HAuCl 4 , CTAB, AgNO 3 , ascorbic acid, hydrochloric acid and sodium borohydride to react to obtain rod-shaped gold nanocrystals; chloroauric acid solution is selected as the gold salt; CTAB is selected as the surfactant; AgNO 3 The concentration of the solution is used as a variable in the training database.

5. The modeling method according to claim 4, It is characterized in that The thermodynamic model curve expression is LSPR = (3.625 × ln (C (Ag + )) 2 -3.43×(C(Ag + ))+0.58×(C(Ag + ))×ln(C(Ag + ))+6.31) ×96+418, where the parameter corresponding to the ratio of the crystal surface energy is the aspect ratio AR=3.625×ln(C(Ag + )) 2 -3.43×(C(Ag + ))+0.58×(C(Ag + ))×ln(C(Ag + ))+6.31; C(Ag + ) represents the silver ion concentration.

6. A prediction method for digital manufacturing of nanocrystals, include: The LSPR value of the nanocrystal to be prepared is obtained, and the corresponding reaction conditions are obtained through the thermodynamic model of digital manufacturing of nanocrystals according to any one of claims 1 to 5.

7. A computer-readable storage medium having a computer program stored thereon, in, When the program is executed by a processor, the steps of the method for modeling a thermodynamic model for digital manufacturing of nanocrystals described in any one of claims 1 to 5 are implemented.

8. A computer device comprising a memory and a processor, wherein a computer program that can be run on the processor is stored in the memory, and when the processor executes the program, the steps of the modeling method of modeling the thermodynamic model of digital manufacturing of nanocrystals as described in any one of claims 1-5 are implemented.

Images