October 2019: Artificial Intelligence reaches Volume Crystal Growth

Artificial inteligence in crystal growthSection Fundamental Discription at IKZ - Achieved accelerated
process development using artificial intelligence (AI)


Crystal growth is essential for the development of new technological functional materials. A special challenge is the reduction of costs and time in the production of industrially important materials. However, a general approach based on trial-and-error experiments and CFD simulations (Computational Fluid Dynamic) is too slow to provide fast answers. For example, increasing the diameter of silicon wafers from 1 inch to 12 inches took 40 years using this method.



However, artificial intelligence (AI) can significantly shorten the development time of crystal growth processes. In order to come closer to this goal, the section "Fundamental Description" of the IKZ has extended its research topics to the investigation of various applications of AI in the growth of volume crystals.

The IKZ used static ANNs (artificial neural networks) for pattern recognition and parameter optimization in the magnetic driven growth of crystalline materials [1,2]. A current research topic is the application of dynamic neural networks for real-time prediction in the transient VGF-GaAs crystal growth process. For example, temperature distributions in the melting furnace as well as the position of the crystallization front during the growth process can be predicted [3]. Such dynamic ANNs enable process automation and control as a decisive step in the development of smart factories in the context of Industry 4.0. The Application of the AI technologies to other crystal growth topics is currently in progress. 

 


[1] N. Dropka, M. Holena, Optimization of magnetically driven directional solidification of silicon using artificial neural networks and Gaussian process models, Journal of Crystal Growth 471 (2017) 53-61.

[2] N. Dropka, M. Holena and Ch. Frank-Rotsch, TMF optimization in VGF crystal growth of GaAs by artificial neural networks and Gaussian process models, Proceedings of XVIII International UIE-Congress on Electrotechnologies for Material Processing, Eds. E. Baake, B. Nacke, Hannover, June 6 - 9, 2017, p.203-208.

[3] N. Dropka, M. Holena, S. Ecklebe, Ch. Frank-Rotsch, J. Winkler, Fast forecasting of VGF crystal growth process by dynamic neural networks, Journal of Crystal Growth 521(2019) 9-14.

 

return to start page

Partnerlogo 1 EN

Partnerlogo 2

Partnerlogo 3