Highly Efficient Global Optimization Method FGO and Its Application in Nanoclusters

Topic: 
Highly Efficient Global Optimization Method FGO and Its Application in Nanoclusters
Date & Time: 
Friday, May 10, 2024 - 09:45 to 10:30
Speaker: 
Liping Chen, Zhejiang Normal University
Location: 
Hosted via Zoom (Meeting ID: 914 8198 6624; Passcode: 461070)

- Join via Zoom -

Abstract:  

Recently, we have proposed a new unbiased global optimization method---Fuzzy Global Optimization (FGO). By calculating and storing the interaction of all possible atom-pair in discrete space, we can achieve fast evaluation of the cluster energy with any structure. Starting from a completely random initial structure, we introduce directed Monte Carlo (DMC) to change the structure and reduce the cluster energy iteratively. Furthermore, we obtain a series of candidate clusters by fine tuning the structures with surface Monte Carlo (SMC). Then the candidate structures are optimized and the exact energies are calculated in real space. Finally, we obtain the putative global minimum (GM) with the lowest energy.

We have applied FGO to successfully investigate Lennard-Jones clusters with up to 1000 atoms, Morse cluster up to 700 atoms and CdSe clusters up to 160 atoms. Compared with other unbiased methods, the investigated cluster sizes have been significantly enlarged in FGO. FGO not only obtains all the structures reported in literatures, but also finds some new GMs with lower energies. FGO can search out different types of structures automatically due to its unbiased character and reveal the growth pattern of clusters and magic sizes. These results indicate that FGO has both high efficiency and accuracy in global optimization, which can be intensively applied to the structural search of different nanoclusters.

Biography:  

Dr. Liping Chen obtained her bachelor and doctor’s degrees from Wuhan University and Institute of Chemistry, Chinese Academy of Sciences (ICCAS), respectively. After her postdoc stay in ICCAS, University of Mons in Belgium and University of Rochester in USA, she joined Zhejiang Normal University in 2019. She involves in the theoretical investigation of complex materials and devices for a long time. Recently, she focuses on the development of highly efficient global optimization method and its application in complex nanoclusters. She has published more than 40 papers in journals such as Nature Communications, AM, PRL, JPCL, JCP and so on.

Seminar Series by the NYU-ECNU Center for Computational Chemistry at NYU Shanghai