Design a New Superconductor from Experiment DataApplication Note Zaptron Systems, Inc.
Goal:
Design a new superconductor with higher Tc (critical temperature) value from
experiment data.
Method:
It was found that when floride (F) is added to the Bi-Ca-Sr-Cu-O superconductor, the Tc of this superconductor can be increased. But a reliable method was not available to guide the material manufacturing process so that less experiments were performed, specially at a time when the heat of the superconductor competition was around the world (1991-1992).
Data mining techniques were used to build a model from past experiment data, and the model was used to extropolate the data space for new superconductor design. The innovation was the ability to separate good data from bad data, so that useful features were selected and a reliable mathematical model could be built.
Result:
The model so obtained from data was used to guide the material design process.
After one extropolation, the Tc was increased to 110ºK to 116ºK, and by a second
extropolation, the Tc was increased to 210ºK. This was a significant improvement at the
time.
© Copyright, 1998-2000, Zaptron Systems, Inc. Back to Home.