【Frontiers in HaiChuan Chemical Technology】South China University of Technology develops an autonomous chemical robotics platform for the end-to-end synthesis of nanoparticles
Thread Content
Chemical autonomous robot platform enables end-to-end synthesis of nanoparticles 2025-08-25 High-quality nanoparticles have significant applications in fields such as energy, catalysis, and medicine. However, AI-powered autonomous exploration of nanomaterials is still constrained by issues such as the high cost of obtaining datasets and poor reproducibility of experiments across platforms. Recently, a team led by Professor Chen Guangxu from South China University of Technology developed a completely new automated experimental system. This platform is built on commercial automation modules, and the GPT model is incorporated to provide researchers with synthetic solutions. The system can directly invoke or edit automated scripts based on the generated experimental steps, thereby enabling the automated execution of the entire experimental process. Meanwhile, by using the A* algorithm for iterative optimization of parameters, this platform has successfully synthesized various nanomaterials such as gold, silver, cuprous oxide, and palladium-copper nanocages, demonstrating significant versatility and potential for further expansion.
Figure 2. Literature mining module and characterization of gold nanobipyramids (Au NBPs). a. Workflow of the literature mining module for nanoparticle synthesis. This module is divided into two sub-modules: the summary module and the query module. The abstract module systematically extracts and processes information from chemical literature through the following steps: document compression, parsing, index construction, and querying. Meanwhile, the query module uses similar methods, including literature retrieval, parsing, index construction, and querying, to identify relevant experimental details and output reagents and procedure sequences. b. UV-Vis spectra of Au NBPs (gold nanotetrahedra) synthesized using optimal parameters. c. Transmission electron microscope (TEM) images of the Au NBPs samples synthesized using optimal parameters. (Scale = 200 nm) d. Statistics of the longitudinal length of the optimal Au NBPs samples. e. Statistical analysis of the radial length of the best Au NBPs samples.
Figure 3. Optimization and synthesis of gold nanorods (Au NR) using the A* algorithm. a. Flowchart of the A* algorithm used for the optimization of Au NRs. The heuristic function of this algorithm adopts the Upper Confidence Bound (UCB) strategy. In the figure, A-D are hyperparameters. b. Normalized UV-visible (UV-vis) spectra of Au NRs with optimized longitudinal surface plasmon resonance (LSPR) peaks (at intervals of 50 nanometers, in the range of 600–900 nanometers). c-i. Transmission electron microscope (TEM) images of the final Au NRs products for different LSPR ranges (scale = 100 nm). j-p. The LSPR values of Au NRs are determined through a parameter search process carried out by the A* algorithm. The SLSPR value is the algorithm’s calculation of the potential value of parameter points prior to experimentation. The ZLSPR value is an evaluation by the algorithm, after experimentation, of how close the parameter point is to the target LSPR.
Figure 4. The A* algorithm optimized with gold nanospheres (Au NSs). a. Schematic diagram of the synthesis process for Au NSs with sizes of 24, 54, 68, and 88 nm. b. The A* algorithm performs two different processes based on the desired diameter of the Au NSs. Evaluation and heuristic functions for optimizing Au NSs: Mean Square Error (MSE) is used to measure the difference in UV-visible (UV-vis) spectra, where Y represents the absorbance of the target spectrum. yi represents the absorbance of the experimental spectrum. n represents the number of spectral points. E is a hyperparameter, set to 50 here for normalization. c. An overview of the optimization process of the A* algorithm for Au NSs, showing the results corresponding to the parameter points throughout the process. d-g. UV-visible spectroscopy curves of Au NSs with target diameter and the finally optimized UV-visible spectroscopy curve. h-k. Diameter distributions of 24, 54, 69, and 88 nm Au NSs under the ultimately optimized parameters of the A* algorithm. i-o. Optimized transmission electron microscope (TEM) images of 24, 54, 68, and 88 nm Au NSs.
【The Defunct Fertilizer Plant】Hefei Jianghuai Fertilizer General Factory, which changed its name over time: https://bbs.hcbbs.com/thread-5699429-1-1.html (Source: Haichuan Chemical Industry Forum)
【Ten Years of Rapid Development in Chemical Engineering Equipment】2635–2025: China’s Self-constructed Largest-tonnage, Highest Gas Storage Capacity New Offshore Floating Natural Gas Liquefaction Facilities https://bbs.hcbbs.com/thread-5699787-1-1.html (Source: Haichuan Chemical Industry Forum)