WO2023164811A1 - Robot scientist-aided crystal material digital manufacturing method, and system - Google Patents

Abstract

A robot scientist-aided crystal material digital manufacturing method and a system. Auxiliary experiments for nanocrystal materials are carried out by means of a robot scientist platform, a database is constructed according to the auxiliary experiments, and then digital manufacturing of the nanocrystal materials is implemented by the robot scientist platform on the basis of the database. High-throughput synthesis of nanocrystal materials having different AR and OD ratios is implemented under the guidance of a machine learning trained model and by using an automated robot HTE platform.

Description

Method and system for robotic scientist-assisted digital fabrication of crystal materials technical field

The invention belongs to the field of new material preparation technology and digital manufacturing technology, and in particular relates to a method, experimental process, database and system for robot scientist-assisted digital manufacturing of crystal materials.

Background technique

Research on the digital and intelligent preparation of materials combined with robotics/process automation and artificial intelligence is in the ascendant at home and abroad. Exploring the cross-integration of 4Paradigm data-driven scientific discovery and materials disciplines will provide a new methodology for the development of new concept materials and materials common science. The cross-integration of material preparation and cutting-edge technologies such as robotics and artificial intelligence has become a development trend and promotes material research and development. The transformation from the traditional mode of "scientific intuition and trial and error" to the new mode of "digitalization and intelligence" can provide greater potential for exploring transformative and subversive new concept materials.

Research on digital manufacturing of robot-assisted materials is not only an international frontier of science and technology, but also a major domestic demand. In October 2021, General Secretary Xi Jinping made an important deployment at the 34th collective meeting of the Political Bureau of the CPC Central Committee: "Promote the deep integration of digital technology and physical industries." This year, the National Natural Science Foundation of China discipline planning and layout: "During the 14th Five-Year Plan period, explore and establish a new research paradigm for intelligent materials." At the same time, the Chinese Academy of Sciences released a list of major scientific issues: "Through the combination of artificial intelligence big data and materials, construct machine scientist".

At the same time, the objects to solve key scientific problems and break through the "stuck neck" technology are becoming increasingly complex and high-dimensional, which brings great challenges to technological innovation. The current mainstream research paradigms such as experiments, theories, and simulations mainly rely on traditional methods such as trial and error and variable dimensionality reduction. It often takes decades to complete complex processes from project approval, pre-research analysis, experiments to industrial trial production. Its limitations and low The effectiveness is becoming more and more obvious. The development of the "Specialized, Specialized and New" industry urgently needs disruptive innovations in research paradigms.

To sum up, the development of digital manufacturing of functional materials and the integration of robotics and artificial intelligence are the frontiers and hotspots of international research. However, it is urgent to open up advanced machine learning algorithms, build a robot scientist platform, and build a material genome database with descriptors. common technical issues.

Patent application CN202111468080.X provides a programmable rational design method for the whole process of nanocrystal material mathematical model, database and AI algorithm, and proposes to build robot-assisted equipment (robot scientist platform) to accelerate digital manufacturing of functional materials, and use artificial intelligence to empower functions Materials accelerate innovation, and the development of machine scientists for performance prediction of functional materials will endow machines with the core of intelligence, enabling them to penetrate into the bottom layer of quantum mechanics with the help of big data and artificial intelligence, refine the structure-effect relationship, and use intelligent algorithms, software engines, and structural and functional information. Inversion provides a database, builds a material genome project based on a machine scientist platform, builds a universal robot scientist platform for research and development of functional materials, realizes reverse synthesis prediction and design of functional materials, and then promotes digital manufacturing of functional materials. For this major task of interdisciplinary integration, the digital manufacturing of functional materials will be accelerated, and the research and development cycle will be greatly shortened to break through the "stuck neck" technology.

However, there have been no similar reports on the application of the robotic scientist platform to the preparation of nanocrystalline materials.

Contents of the invention

The purpose of the present invention is to provide a nano crystal material manufacturing method, system, computer equipment and storage medium based on the robot scientist platform, utilize the automatic robot HTE (High Throuput Experiment) platform, under the guidance of the machine learning training model, through the seedless Growth method to enable digital fabrication of nanocrystalline materials.

One aspect of the present invention provides a method for robotic scientist-assisted digital fabrication of crystal materials, comprising the following steps:

Auxiliary experiments with nanocrystalline materials via robotic scientist platforms;

Build a database according to the auxiliary experiments;

Based on the database, the robotic scientist platform enables digital fabrication of nanocrystalline materials.

In the steps of performing auxiliary experiments on nanocrystalline materials through the robot scientist platform, specifically include:

Orthogonal experiments on nanocrystalline materials are performed through the robotic scientist platform.

In some of these embodiments, in the step of carrying out the orthogonal experiment of the nanocrystal material through the robot scientist platform, the following steps are specifically included:

An orthogonal array containing reactant concentrations (proportions) is input into the robot scientist platform;

Select the reactant concentration (ratio) as a single feature;

High-throughput synthesis of nanocrystalline materials is performed in the robot scientist platform based on the single feature.

In some of these embodiments, in the step of performing the orthogonal experiment of the nanocrystal material through the robot scientist platform, the following steps are also included:

The robot scientist platform repeatedly executes the steps of high-throughput synthesis of nanocrystal materials according to the ultraviolet-visible-near-infrared absorption spectrum of the synthesized nanocrystal materials.

In some of these embodiments, in the step of performing the orthogonal experiment of the nanocrystal material through the robot scientist platform, the following step is further included: the robot scientist platform repeatedly performs high Steps for throughput synthesis of nanocrystalline materials.

In some of these embodiments, in the step of performing the orthogonal experiment of the nanocrystal material through the robot scientist platform, the following step is further included: the robot scientist platform repeatedly performs high-pass experiments according to the optical density of the synthesized nanocrystal material. Quantitative steps in the synthesis of nanocrystalline materials.

In some of the embodiments, in the step of constructing the database according to the auxiliary experiment, the database includes reactant concentration (ratio), ultraviolet-visible-near-infrared absorption spectrum, aspect ratio and optical density.

In some of these embodiments, based on the database, the robotic scientist platform realizes digital manufacturing of nanocrystalline materials, specifically including the following steps:

Relying on machine learning algorithms, the robotic scientist platform enables digital fabrication of nanocrystalline materials.

In some of the embodiments, a step of further verifying the prepared nanocrystals by TEM and color features is also included.

Another aspect of the present invention provides a system for robot scientists to assist digital manufacturing of crystal materials, including: the robot scientist platform, which includes an auxiliary experiment module, a database construction module and a digital manufacturing module,

The auxiliary experiment module carries out the automated auxiliary experiment of nanocrystal material synthesis through the robot scientist platform; the database construction module constructs a database according to the automated experiment; the digital manufacturing module is based on the database, and the robot scientist platform realizes the nanometer Digital fabrication of crystalline materials.

In some of these embodiments, the robotic scientist platform conducts orthogonal experiments on nanocrystalline materials.

In some of these embodiments, the robotic scientist platform conducts orthogonal experiments on nanocrystalline materials, specifically including:

Input an orthogonal experimental design comprising reactant concentrations (ratio) on the robot scientist platform;

Select the concentration (ratio) of the reactant;

High-throughput synthesis of nanocrystalline materials is performed in the robotic scientist platform according to the reactant concentrations or ratios.

In some of these embodiments, in the orthogonal experiment of nanocrystalline materials carried out by the robot scientist platform, it also includes:

The robotic scientist platform repeatedly performs high-throughput synthesis of nanocrystalline materials based on their UV-Vis-NIR absorption spectra.

In some of these embodiments, in performing the orthogonal experiments on nanocrystal materials by the robot scientist platform, it also includes: the robot scientist platform repeatedly executes high-throughput synthesis of nanocrystals according to the aspect ratio of the synthesized nanocrystal materials Material.

In some of the embodiments, in the orthogonal experiment of nanocrystal materials performed by the robot scientist platform, it also includes: the robot scientist platform repeatedly executes high-throughput synthesis of nanocrystal materials according to the optical density of the synthesized nanocrystal materials .

In some of the embodiments, the database includes reactant concentration (ratio), concentration (ratio), ultraviolet-visible-near-infrared absorption spectrum, aspect ratio and optical density.

In some of these embodiments, based on the database, the robot scientist platform realizes digital manufacturing of nanocrystal materials, specifically including:

Relying on machine learning algorithms, the robotic scientist platform enables digital fabrication of nanocrystalline materials.

In some of the embodiments, the prepared nanocrystals are further verified by TEM and color characteristics.

The third aspect of the present invention provides a computer device. The computer device includes a processor and a memory connected to the processor. Program instructions are stored in the memory. When the program instructions are executed by the processor, , causing the processor to execute the steps of the method for robotic scientist-assisted digital fabrication of crystal materials.

The fourth aspect of the present invention provides a storage medium storing program instructions capable of implementing the method for robot-scientist-assisted digital manufacturing of crystal materials.

The application adopts the above-mentioned technical solution, which has the following beneficial effects:

The present invention provides a nanocrystal material manufacturing method, system, computer equipment, and storage medium based on a robot scientist platform. The robot scientist platform is used to conduct orthogonal experiments on nanocrystal materials, and a database is constructed according to the auxiliary experiments, and then based on The database and the robot scientist platform realize the digital manufacturing of nanocrystal materials. This application utilizes the automated robot HTE platform, under the guidance of the machine learning generation model, through seedless growth, to realize the nanocrystal materials with different AR and OD ratios High-throughput synthesis accelerates digital manufacturing of functional materials, improves manufacturing efficiency of crystal materials and shortens R&D cycle.

Description of drawings

FIG. 1 is a schematic diagram of the principles of the nanocrystal material digital manufacturing method and system provided by the present application on the robot scientist platform.

Fig. 2 is a flow chart of the steps of the method for robot scientist-assisted digital manufacturing of crystal materials provided by the present application.

3 is a schematic structural diagram of anisotropic gold nanorods with different localized surface plasmon resonance (LSPR) peaks prepared by using an orthogonal factorial experimental design in Example 1 of the present application.

Fig. 4 is a schematic structural diagram of a system for robotic scientist-assisted digital manufacturing of crystal materials provided in Embodiment 2 of the present application.

Detailed ways

In order to make the above objects, features and advantages of the present invention more obvious and understandable, the specific implementation modes of the present invention will be described in detail below, but they should not be construed as limiting the scope of implementation of the present invention.

Please refer to FIG. 1 , which is a schematic diagram of the principle of the nanocrystal material digital manufacturing method and system on the robot scientist platform provided by the present application.

This application uses the robotic scientist platform (HTE platform) to conduct an orthogonal experimental design (a in Figure 2) to study the selectivity of a single feature (Single Feature, SF), through the concentration change of each reactant (step a), obtained The experimental data is used to discover the structure-activity relationship between the structure modulating predictors (Structure Modulating Precursors, SMP) and AR, OD, input into the machine learning algorithm (step c), and finally guide the double feature (Double Feature, DF) and triple feature (Triple Feature ,TF) experiment execution and machine learning model establishment, the purpose is to expand the aspect ratio range of anisotropic nanorods and optimize the ratio of optical density, extract key information such as LSPR, OD value and so on from its absorption spectrum, and pass the color The map was visualized (step d), and further verified by TEM and color features (step e), finally leading to high-throughput synthesis of nanocrystalline materials. For the convenience of description, the technical solution of the present application will be described in detail below by taking the preparation of AuNRs as an example.

Example 1

As shown in Fig. 2, a flow chart of the steps of the method for robotic scientist-assisted digital fabrication of crystal materials is provided, including the following steps:

Step S110: Carry out auxiliary experiments on nanocrystalline materials through the robot scientist platform.

In this embodiment, the robot scientist platform comes from the applicant's patent application CN202111468080.X, and its detailed technical solution has been described in detail in the aforementioned application, so it will not be repeated here.

In some of these embodiments, the robotic scientist platform conducts orthogonal experiments on nanocrystalline materials.

In this embodiment, the orthogonal experiment of nanocrystal materials is carried out through the robot scientist platform, which specifically includes the following steps:

Step S111: Inputting an orthogonal array containing reactant concentrations (proportions) on the robot scientist platform.

In this example, the preparation of AuNRs with different ARs is taken as an example for illustration. In the preparation process, the selection of the amount, time, temperature, and experimental system (total volume of the reaction) included in the reaction is suitable for the robot scientist platform. Taking this into account, the input contains an orthogonal array of reactant concentrations (proportions), And by generating sufficient numbers, making reasonable comparisons, and expanding the range of subsequent databases, the model has higher accuracy and increased applicability. Referring to the table below, this example provides an orthogonal factorial model for several different reactant ratios/concentrations.

Table 1 - Experimental fabrication of gold nanorods using an orthogonal design

Step S112: Select the concentration (ratio) of the reactant therein.

Referring again to Figure 1, this example provides an orthogonal factorial model for different reactant ratios/concentrations, although simultaneous triplicate determinations of reproducibility follow a similar default pattern of mixing time and sequence of precursor addition, regardless of concentration and volume The possibility of variable (CTAB, HAuCl ₄ , AgNO ₃ , HCl, NaBH ₄ , AA) robotic scientist platforms to repeat experiments in shorter time intervals also eliminates the reproducibility challenges posed by miniaturization and parallelization. Therefore, an orthogonal design was used to determine the optimal formulation needed to determine SMP in further experiments.

Please refer to Table 2, which represents the non-nucleated growth of Au nanorods with different 24 CTAB concentration levels in a single feature experimental setup, and the other 6 features kept the default concentrations unchanged.

Experiment number	CTAB concentration	TSPR	OD(T)	LSPR	OD(L)	OD Ratio		Aspect ratio
1	0	547	0.04	-	-	-	-
2	0.01	529	0.14	-	-	-	-
3	0.02	533	0.58	-	-	-	-
4	0.03	547	0.88	-	-	-	-
5	0.04	524	0.79	699	0.84	1.06	2.94
6	0.05	520	0.78	715	0.93	1.19	3.11
7	0.06	515	0.71	742	1.05	1.48	3.39
8	0.07	513	0.70	757	1.25	1.79	3.55
9	0.08	512	0.66	759	1.24	1.88	3.57
10	0.09	510	0.65	757	1.37	2.11	3.55
11	0.1	510	0.67	761	1.53	2.28	3.59
12	0.11	510	0.67	772	1.57	2.34	3.71
13	0.12	512	0.68	733	1.45	2.13	3.29

14	0.13	517	0.72	724	1.21	1.68	3.20
15	0.14	521	0.71	724	1.14	1.61	3.20
16	0.15	517	0.73	720	1.23	1.68	3.16
17	0.16	521	0.73	717	1.05	1.44	3.13
18	0.17	533	0.83	-	-	-	-
19	0.18	536	0.82	-	-	-	-
20	0.19	533	0.87	-	-	-	-
twenty one	0.2	536	0.88	-	-	-	-
twenty two	0.21	536	0.91	-	-	-	-
twenty three	0.22	533	0.93	-	-	-	-
twenty four	0.23	533	0.97	-	-	-	-

Please refer to Table 3, which represents the non-nucleated growth of Au nanorods with different 24 HCL concentration levels in a single feature experimental setup, and the other 6 features kept the default concentrations unchanged.

Experiment number	HCL molar concentration	TSPR	OD(T)	LSPR	OD(L)	OD Ratio		Aspect ratio
1	0	522	0.92	721	1.04	1.13	3.17
2	0.1	523	0.88	729	0.99	1.13	3.25
3	0.2	523	0.89	732	0.93	1.04	3.28
4	0.3	525	0.85	746	0.89	1.05	3.43
5	0.4	526	0.84	749	0.93	1.11	3.46
6	0.5	521	0.77	759	1.03	1.34	3.57
7	0.6	517	0.70	769	1.21	1.73	3.67
8	0.7	512	0.68	772	1.38	2.03	3.71
9	0.8	512	0.66	788	1.38	2.09	3.87
10	0.9	510	0.63	789	1.54	2.44	3.88
11	1	510	0.61	811	1.58	2.59	4.12
12	1.1	510	0.62	811	1.63	2.63	4.12
13	1.2	508	0.65	795	1.80	2.77	3.95
14	1.3	508	0.63	785	1.82	2.89	3.84
15	1.4	509	0.62	793	1.80	2.90	3.93
16	1.5	508	0.62	801	1.78	2.87	4.01
17	1.6	508	0.61	805	1.81	2.97	4.05
18	1.7	508	0.61	795	1.89	3.10	3.95
19	1.8	507	0.59	807	1.83	3.10	4.07
20	1.9	508	0.58	807	1.78	3.07	4.07
twenty one	2	508	0.57	828	1.91	3.35	4.29
twenty two	2.1	508	0.57	820	1.88	3.30	4.21
twenty three	2.2	508	0.54	825	1.78	3.30	4.26
twenty four	2.3	508	0.55	840	1.83	3.33	4.42

It can be understood that in practice, the volume or concentration of HAuCl ₄ , AgNO ₃ , NaBH ₄ , NaOL, and AA can also be adjusted, while the default concentrations of the other six functions remain unchanged, which will not be described here.

Step S113: Perform high-throughput synthesis of nanocrystal materials in the robot scientist platform according to the reactant concentration or ratio.

Please refer to FIG. 3 , which is a schematic diagram of the structure of anisotropic gold nanorods with different LSPR peaks prepared by using an orthogonal factor model in this embodiment. Among them, (a) in the figure represents the image captured after 2 hours of nucleation, and each experiment was repeated 3 times. Figures 2b-d show the absorption spectra of some orthogonal experimental products, indicating the positions of TSPR (transverse resonance peak) and LSPR (radial resonance peak) corresponding to the ultraviolet-visible-near-infrared region, respectively. Figure 3c–d demonstrates the reliability and reproducibility of the experiment. e–h in Fig. 3 represent the ratios of AR and OD obtained with different reactant concentrations.

In some of these embodiments, in the step of performing the orthogonal experiment of the nanocrystal material through the robot scientist platform, the following steps are also included:

Step S114: The robot scientist platform repeatedly executes the steps of high-throughput nanocrystal material synthesis according to the ultraviolet-visible-near-infrared absorption spectrum of the synthesized nanocrystal material.

It can be understood that after the high-throughput synthesis of nanocrystal materials in the robot scientist platform, the detection of the ultraviolet-visible light-near-infrared absorption spectrum can be performed, and the concentration (ratio) of the reactants can be adjusted according to the detection situation, and the high-throughput Steps to synthesize nanocrystalline materials for more optimized product performance and digital fabrication of nanocrystalline materials.

It can be appreciated that a dual-feature experimental design was employed in this example to obtain 96 different Au nanorods by a seedless method by combining different concentration levels (numbers) of CTAB (indicated by C) and HCl (indicated by H) Growth colloid, the other 5 features were kept at the default concentrations previously used in single feature experiments, see the following list for details.

Further, grade 96 results were obtained from the seedless growth of Au nanorods by the dual-feature technique with different concentrations of CTAB and HCl pairs, and the other 5 features were kept at the default concentrations previously used in single-feature experiments, as detailed in the table below.

Experiment number	TSPR	OD(T)	LSPR	OD(L)	OD Ratio		Aspect ratio
1	534	0.97	-	-	-	-
2	534	0.95	-	-	-	-
3	518	0.82	715	1.05	1.28	3.11
4	518	0.79	726	1.10	1.39	3.22
5	515	0.75	754	1.19	1.59	3.52
6	518	0.75	752	1.12	1.48	3.49
7	524	0.80	734	1.02	1.28	3.31
8	525	0.78	750	1.02	1.30	3.47

9	518	0.74	753	1.10	1.49	3.51
10	511	0.65	783	1.35	2.07	3.82
11	508	0.61	809	1.70	2.77	4.09
12	509	0.60	777	1.53	2.56	3.76
13	529	0.92	691	0.85	0.92	2.85
14	528	0.91	702	0.84	0.92	2.97
15	526	0.83	741	0.90	1.08	3.38
16	525	0.78	729	0.93	1.19	3.25
17	523	0.76	749	1.03	1.35	3.46
18	518	0.72	754	1.09	1.51	3.52
19	515	0.69	771	1.26	1.81	3.69
20	515	0.67	794	1.23	1.84	3.94
twenty one	518	0.74	758	1.09	1.49	3.56
twenty two	511	0.65	758	1.33	2.04	3.56
twenty three	507	0.59	808	1.63	2.75	4.08
twenty four	510	0.60	771	1.55	2.59	3.69
25	528	0.90	710	0.86	0.96	3.05
26	525	0.80	750	0.97	1.21	3.47
27	518	0.74	726	0.96	1.30	3.22
28	515	0.71	732	1.00	1.42	3.28
29	515	0.68	741	1.08	1.59	3.38
30	511	0.64	806	1.28	2.01	4.06
31	511	0.65	760	1.20	1.84	3.58
32	511	0.64	765	1.28	2.02	3.63
33	518	0.73	751	1.04	1.42	3.48
34	513	0.64	765	1.16	1.82	3.63
35	509	0.60	770	1.53	2.53	3.68

36	510	0.59	783	1.42	2.42	3.82
37	525	0.88	729	0.91	1.03	3.25
38	521	0.80	731	0.94	1.18	3.27
39	515	0.73	741	1.00	1.38	3.38
40	515	0.70	734	1.05	1.49	3.31
41	512	0.67	747	1.15	1.72	3.44
42	511	0.65	757	1.24	1.91	3.55
43	511	0.64	765	1.32	2.05	3.63
44	511	0.63	769	1.36	2.18	3.67
45	518	0.75	741	1.04	1.40	3.38
46	511	0.64	750	1.28	1.99	3.47
47	509	0.59	799	1.65	2.80	3.99
48	510	0.59	780	1.51	2.55	3.79
49	523	0.84	741	0.92	1.09	3.38
50	518	0.76	741	1.02	1.36	3.38
51	515	0.69	769	1.05	1.52	3.67
52	513	0.67	741	1.08	1.61	3.38
53	511	0.65	750	1.24	1.89	3.47
54	510	0.62	775	1.37	2.21	3.74
55	510	0.62	779	1.40	2.25	3.78
56	510	0.61	790	1.41	2.31	3.89
57	511	0.60	791	1.56	2.58	3.91
58	509	0.56	808	1.73	3.08	4.08
59	509	0.55	831	1.73	3.13	4.33
60	509	0.57	822	1.78	3.10	4.23
61	522	0.80	749	1.03	1.29	3.46
62	515	0.73	753	1.11	1.51	3.51

63	513	0.68	760	1.14	1.69	3.58
64	511	0.65	762	1.28	1.96	3.60
65	511	0.63	765	1.39	2.21	3.63
66	510	0.61	783	1.51	2.48	3.82
67	510	0.61	790	1.53	2.50	3.89
68	508	0.58	805	1.63	2.79	4.05
69	510	0.61	790	1.62	2.68	3.89
70	510	0.56	806	1.67	2.97	4.06
71	509	0.54	822	1.74	3.20	4.23
72	510	0.56	831	1.77	3.18	4.33
73	518	0.75	753	1.14	1.51	3.51
74	515	0.70	752	1.19	1.70	3.49
75	511	0.66	775	1.30	1.97	3.74
76	511	0.64	757	1.44	2.24	3.55
77	511	0.61	799	1.54	2.54	3.99
78	509	0.59	797	1.65	2.80	3.97
79	510	0.58	808	1.75	3.00	4.08
80	510	0.58	817	1.74	3.02	4.18
81	510	0.60	799	1.78	2.96	3.99
82	511	0.56	822	1.73	3.08	4.23
83	509	0.55	848	1.89	3.41	4.51
84	511	0.55	844	1.84	3.36	4.46
85	518	0.78	742	1.07	1.38	3.39
86	513	0.70	765	1.28	1.84	3.63
87	511	0.65	765	1.37	2.10	3.63
88	510	0.62	775	1.49	2.41	3.74
89	509	0.61	787	1.63	2.68	3.86

90	509	0.58	821	1.85	3.19	4.22
91	508	0.58	822	1.92	3.30	4.23
92	507	0.57	846	1.90	3.34	4.48
93	510	0.60	803	1.75	2.89	4.03
94	511	0.57	822	1.79	3.12	4.23
95	511	0.52	844	1.76	3.41	4.46
96	511	0.56	837	1.68	3.02	4.39

It can be understood that other dual-feature combinations can be used in practice while the default concentrations of the other five functions remain unchanged, which will not be described here.

In some of these embodiments, in the step of performing the orthogonal experiment of the nanocrystal material through the robot scientist platform, the following steps are also included:

Step S115: The robotic scientist platform repeatedly executes the step of high-throughput synthesis of nanocrystal materials according to the aspect ratio of the synthesized nanocrystal materials.

It can be understood that the aspect ratio can be detected after the high-throughput synthesis of nanocrystal materials is performed in the robot scientist platform, and the concentration (ratio) of reactants is adjusted according to the detection situation, and the steps of high-throughput synthesis of nanocrystal materials are repeated , to achieve more optimized product performance.

It can be understood that in this embodiment, a three-characteristic experimental design was adopted to obtain ₆₄ different colloids for the growth of gold nanorods by using a three-characteristic experimental design. and CTAB (indicated by C). The other 4 features were kept at the default concentrations previously used in single feature experiments, see the list below for details.

Further, 64-level results were obtained from the seedless growth of Au nanorods by the triple-feature technique with different concentrations of _HAuCl4 , HCl, and CTAB, and the other 4 features were kept at the default concentrations previously used in single-feature experiments. See the table below for details.

Example	TSPR	OD(T)	LSPR	OD(L)	OD Ratio		Aspect ratio
1	529	0.62	-	-	-	-
2	525	0.54	-	-	-	-
3	524	0.51	-	-	-	-
4	525	0.51	-	-	-	-
5	528	0.47	716	0.39	0.83	3.12
6	515	0.38	780	0.54	1.42	3.79
7	515	0.36	752	0.53	1.45	3.49
8	515	0.35	791	0.69	1.98	3.91
9	532	1.13	-	-	-	-
10	523	0.91	716	0.92	1.01	3.12
11	519	0.86	712	1.00	1.16	3.07
12	536	0.80	718	1.04	1.29	3.14
13	530	0.84	728	0.82	0.98	3.24
14	513	0.66	795	1.26	1.91	3.95
15	509	0.61	803	1.52	2.49	4.03
16	511	0.62	750	1.47	2.36	3.47
17	-	-	-	-	-	-
18	523	1.27	693	1.37	1.09	2.87

19	521	1.19	708	1.51	1.27	3.03
20	521	1.19	695	1.47	1.24	2.89
twenty one	523	1.16	696	1.34	1.15	2.91
twenty two	511	0.97	752	2.00	2.07	3.49
twenty three	511	0.89	756	1.98	2.21	3.54
twenty four	511	0.80	749	1.69	2.11	3.46
25	535	2.13	-	-	-	-
26	523	1.58	698	2.08	1.31	2.93
27	523	1.54	697	2.14	1.39	2.92
28	523	1.56	696	2.08	1.33	2.91
29	521	1.47	701	2.15	1.46	2.96
30	513	1.26	751	2.64	2.10	3.48
31	513	1.15	735	2.31	2.01	3.32
32	515	1.04	713	1.89	1.82	3.08
33	521	0.40	765	0.41	1.03	3.63
34	511	0.34	810	0.70	2.06	4.11
35	515	0.33	790	0.75	2.25	3.89
36	515	0.28	800	0.58	2.02	4.00
37	515	0.36	794	0.48	1.33	3.94
38	515	0.31	831	0.68	2.22	4.33
39	521	0.25	841	0.50	2.03	4.43
40	522	0.15	794	0.25	1.60	3.94
41	515	0.74	740	1.17	1.59	3.37
42	509	0.63	794	1.82	2.90	3.94
43	509	0.59	795	1.81	3.05	3.95
44	511	0.56	780	1.55	2.80	3.79
45	511	0.65	771	1.27	1.95	3.69

46	507	0.55	832	1.74	3.14	4.34
47	511	0.46	810	1.35	2.97	4.11
48	515	0.28	779	0.62	2.17	3.78
49	515	1.06	727	1.90	1.78	3.23
50	509	0.94	774	2.71	2.89	3.73
51	511	0.89	768	2.38	2.66	3.66
52	512	0.83	746	1.88	2.26	3.43
53	511	0.94	765	2.22	2.35	3.63
54	509	0.81	783	2.30	2.86	3.82
55	512	0.72	757	1.62	2.25	3.55
56	515	0.54	726	0.99	1.83	3.22
57	513	1.46	720	2.61	1.79	3.16
58	512	1.25	760	3.20	2.57	3.58
59	510	1.18	739	2.65	2.25	3.36
60	513	1.12	711	2.07	1.86	3.06
61	511	1.27	752	2.75	2.17	3.49
62	511	1.08	743	2.59	2.40	3.40
63	513	0.97	719	1.90	1.95	3.15
64	518	0.72	693	1.05	1.46	2.87

It can be understood that in practice, other combinations of the three features can be used while the default concentrations of the other four functions remain unchanged, which will not be described here.

In some of these embodiments, in the step of carrying out the orthogonal array assisted experiment of nanocrystalline materials through the robot scientist platform, the following steps are also included:

Step S116: The robot scientist platform repeatedly executes the step of high-throughput synthesis of nanocrystal materials according to the optical density of the synthesized nanocrystal materials.

It can be understood that the detection of optical density can be performed after the high-throughput synthesis of nanocrystal materials is performed in the robot scientist platform, and the concentration (ratio) of reactants is adjusted according to the detection situation, and the steps of high-throughput synthesis of nanocrystal materials are repeated, In order to achieve better product performance.

Step S120: constructing a database according to the auxiliary experiment.

It can be understood that after the synthesis of high-throughput synthetic nanocrystal materials is performed in the robot scientist platform, the synthesized products can be detected including ultraviolet-visible-near-infrared absorption spectrum, aspect ratio and optical density, and according to the detection According to the situation, adjust the reactant concentration (ratio) in real time to achieve the best product performance, and save the reaction conditions in the subsequent database.

In some of these embodiments, the database includes reactant concentrations (ratio), ultraviolet-visible-near-infrared absorption spectra, aspect ratios, and optical densities, and may also include other conditions of the reaction, for example, including temperature, reaction time Choice of microscale, etc.

Step S130: Based on the database, the robotic scientist platform implements digital fabrication of nanocrystalline materials.

In some of these embodiments, based on the database, the robotic scientist platform realizes digital manufacturing of nanocrystalline materials, specifically including the following steps:

Step S131: Relying on machine learning algorithms, the robotic scientist platform realizes digital fabrication of nanocrystalline materials.

It can be understood that modeling based on machine learning algorithms can determine the most accurate equation (model) to describe the selected target characteristics (such as aspect ratio and optical density) and the data required for the main characteristics of overall correlation The relationship between the sets; further, for model validation, another algorithm using the scikit-learn package as an alternative is used to train the machine learning model, with the purpose of predicting the aspect ratio of various data sets beyond the experimental range and optical density.

In some of the embodiments, after the digital fabrication of the nanocrystal material is completed, a step of further characterizing the prepared nanocrystal by TEM is also included.

In this embodiment, the image in the growing crystal material is captured by the above robot scientist platform, and then transferred to a spectrophotometer to measure the absorption spectrum in the ultraviolet-visible-near-infrared region, using a UV-Vis-NIR spectrophotometer (Multiskan Skyhigh microplate reader) obtained the absorption spectrum of the 350-1000 nm solution. The morphology and structure were characterized by transmission electron microscope FEI Tecnai G2.

In the above-mentioned embodiment 1 of the present application, the orthogonal experimental design is carried out through the robot scientist platform, and the single-feature selective research experiment is first carried out. Through the concentration change of each reactant, the corresponding relationship between the obtained data reactant concentration and AR, OD, The input machine learning algorithm ultimately guides the execution of dual- and triple-feature experiments with the aim of optimizing aspect ratio and optical density ratio, decoded from their absorption spectrum and visualized by color map, which is further analyzed by TEM and color features To verify and finally obtain the high-throughput synthesis of nanocrystalline materials, this application uses the automated robot HTE platform, under the guidance of machine learning generation models, to achieve high-throughput nanocrystalline materials with different AR and OD ratios through seedless growth Synthesis, accelerating digital manufacturing of functional materials, improving manufacturing efficiency and R&D cycle.

Example 2

Please refer to FIG. 4 , another aspect of the present invention provides a system for robot scientists to assist digital manufacturing of crystal materials, including: the robot scientist platform includes an auxiliary experiment module 110 , a database construction module 120 and a digital manufacturing module 130 .

The auxiliary experiment module 110 performs automated auxiliary experiments on nanocrystal material synthesis through the robot scientist platform. The auxiliary experiment module 110 includes a data input unit 111 , a reactant concentration or ratio selection unit 112 and a high-throughput synthesis unit 113 .

The data input unit 111, input on the robotic scientist platform includes an orthogonal array of reactant concentrations or ratios.

The reactant concentration or ratio selection unit 112 is used to select the reactant concentration or ratio therein.

The high-throughput synthesis unit 113 is used to perform high-throughput synthesis of nanocrystalline materials in the robotic scientist platform according to the reactant concentrations or ratios.

The database construction module 120 constructs a database according to the automated experiment.

The digital fabrication module 130 is based on the database, and the robotic scientist platform enables digital fabrication of nanocrystalline materials.

Its detailed implementation scheme has been described in detail in Embodiment 1, and will not be repeated here.

It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. For the same and similar parts in each embodiment, refer to each other, that is, Can. As for the device-type embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and for related parts, please refer to part of the description of the method embodiments.

In the above-mentioned embodiment 2 of this application, the orthogonal experimental design is carried out through the robot scientist platform, prior to the selectivity of the single-feature technology, through the concentration change of each reactant, the data obtained help simplify and combine the concentration of the reactant, and input the machine Learning algorithms, ultimately guiding the execution of dual- and triple-feature techniques, aim to optimize noise-free aspect and optical density ratios, decoded from their absorption spectra, and visualized by colormaps, which are further analyzed by TEM and color features. For further verification, the high-throughput synthesis of nanocrystalline materials is finally obtained. This application utilizes the automated robot HTE platform, under the guidance of the machine learning generation model, through seedless growth, to achieve high yields of nanocrystalline materials with different AR and OD ratios. Throughput synthesis, accelerate digital manufacturing of functional materials, improve manufacturing efficiency and R&D cycle.

The above is only the implementation mode of this application, and does not limit the scope of patents of this application. Any equivalent structure or equivalent process conversion made by using the contents of this application specification and drawings, or directly or indirectly used in other related technical fields, All are included in the scope of patent protection of the present application in the same way.

Claims (18)

1. A method for robotic scientist-assisted digital manufacturing of crystal materials, characterized in that it comprises the following steps:

   Carry out automated auxiliary experiments for nanocrystal material synthesis through the robot scientist platform;

   Constructing a database according to the automated experiments;

   Based on the database, the robotic scientist platform enables digital fabrication of nanocrystalline materials.
2. The method for robot scientist-assisted digital manufacturing of crystalline materials according to claim 1, characterized in that, in the step of assisting the automated experiment of nano-crystalline materials through the robot scientist platform, specifically comprising:

   Orthogonal experimental design and automated experiments for nanocrystalline materials are performed through the robot scientist platform.
3. The method for robot scientist-assisted digital manufacturing of crystalline materials according to claim 2, wherein, in the step of performing automated experiments on nanocrystalline materials through the robot scientist platform, specifically comprising the following steps:

   Input to the robot scientist platform includes an orthogonal array of reactant concentrations or ratios;

   Choose the reactant concentration or ratio;

   High-throughput synthesis of nanocrystalline materials is performed in the robotic scientist platform according to the reactant concentrations or ratios.
4. The method for robot-scientist-assisted digital manufacturing of crystal materials according to claim 3, characterized in that, in the step of carrying out the orthogonal experiment of nano-crystal materials through the robot scientist platform, the following steps are also included:

   The robot scientist platform repeatedly executes the steps of high-throughput synthesis of nanocrystal materials according to the ultraviolet-visible-near-infrared absorption spectrum of the synthesized nanocrystal materials.
5. The method for robot scientist-assisted digital manufacturing of crystalline materials according to claim 3, characterized in that, in the step of carrying out the orthogonal experiment of nano-crystal materials through the robot scientist platform, the following steps are also included: the robot scientist The platform repeatedly performs the steps of high-throughput synthesis of nanocrystalline materials according to the aspect ratio of the synthesized nanocrystalline materials.
6. The method for robot scientist-assisted digital manufacturing of crystalline materials according to claim 3, characterized in that, in the step of carrying out the orthogonal experiment of nano-crystal materials through the robot scientist platform, the following steps are also included: the robot scientist The platform repeatedly performs the steps of high-throughput synthesis of nanocrystalline materials according to the optical density of the synthesized nanocrystalline materials.
7. The method for robot scientist-assisted digital manufacturing of crystal materials according to claim 2 or 3 or 4 or 5 or 6, characterized in that, in the step of building a database according to the auxiliary experiment, the database includes reactant concentrations or ratios , UV-visible-near-infrared absorption spectrum, nano-gold aspect ratio and optical density.
8. The method for robot scientist-assisted digital manufacturing of crystalline materials according to claim 7, characterized in that, based on the database, the robot scientist platform realizes the digital manufacturing of nano-crystalline materials, specifically comprising the following steps:

   Relying on machine learning algorithms, the robotic scientist platform enables digital fabrication of nanocrystalline materials.
9. The method for digital manufacturing of crystal materials assisted by robot scientists according to claim 1, further comprising the step of further verifying the prepared nanocrystals through color characteristics and transmission electron microscopy.
10. A system for robotic scientist-assisted digital fabrication of crystal materials, characterized by comprising:

    Robot scientist platform, said robot scientist platform includes an auxiliary experiment module, a database building module and a digital manufacturing module,

    The auxiliary experiment module carries out the automated auxiliary experiment of nanocrystal material synthesis through the robot scientist platform; the database construction module constructs a database according to the automated experiment; the digital manufacturing module is based on the database, and the robot scientist platform realizes the nanometer Digital fabrication of crystalline materials.
11. The system for digital manufacturing of crystalline materials assisted by robot scientists according to claim 10, characterized in that the robot scientist platform conducts orthogonal experiments on nanocrystalline materials.
12. The system for digital manufacturing of crystalline materials assisted by robotic scientists according to claim 10, wherein the robotic scientist platform performs orthogonal experiments on nanocrystalline materials, specifically comprising:

    Inputting an orthogonal experimental design including reactant concentrations or ratios on the robot scientist platform;

    Choose the reactant concentration or ratio;

    High-throughput synthesis of nanocrystalline materials is performed in the robotic scientist platform according to the reactant concentrations or ratios.
13. The system for robot scientist-assisted digital manufacturing of crystal materials according to claim 11, characterized in that, in the orthogonal experiment of nano-crystal materials carried out through the robot scientist platform, it also includes:

    The robotic scientist platform repeatedly performs high-throughput synthesis of nanocrystalline materials based on the UV-Vis-NIR absorption spectrum of the synthesized nanocrystalline materials.
14. The system for robot scientist-assisted digital manufacturing of crystal materials according to claim 11, characterized in that, in performing the orthogonal experiment on nanocrystal materials through the robot scientist platform, it also includes: the robot scientist platform according to the synthesized nanometer Aspect Ratio Characterization of Crystalline Materials Reproducibly perform high-throughput synthesis of nanocrystalline materials.
15. The system for robot scientist-assisted digital manufacturing of crystal materials according to claim 11, characterized in that, in performing the orthogonal experiment on nanocrystal materials through the robot scientist platform, it also includes: the robot scientist platform according to the synthesized nanometer Densitometric Characterization of Crystalline Materials Repeatedly perform high-throughput synthesis of nanocrystalline materials.
16. According to claim 10 or 11 or 12 or 13 or 14, the system for assisting digital manufacture of crystal materials by robot scientists is characterized in that the database includes reactant concentrations or ratios, ultraviolet-visible light-near-infrared absorption spectra, long-diameter compared to optical density.
17. The system for robot scientist-assisted digital manufacturing of crystal materials according to claim 10, wherein, based on the database, the robot scientist platform realizes digital manufacturing of nano-crystal materials, specifically comprising:

    Relying on machine learning algorithms, the robotic scientist platform enables digital fabrication of nanocrystalline materials.
18. The system for robot scientist-assisted digital manufacturing of crystal materials according to claim 10, further comprising a step of further verifying the prepared nanocrystals through TEM and color characteristics.

Images