Presentation Details
| Determining Optical Properties for TEC-12D and CdSe(y)Te(1-y) Alloys for Machine Learning Enhanced Spectroscopic Ellipsometry Alex V.Bordovalos1, 2, Nadeesha P.Nadeesha1, 2, Marie S.Tumusange1, 2, Bishal Shrestha1, 2, Suresh Chaulagain1, 2, Randy J.Ellingson1, 2, Nikolas J.Podraza1, 2. 1University of Toledo Department of Physics and Astronomy, Toledo, OH, USA.2Wright Center for Photovoltaic Innovation and Commercialization, University of Toledo, Toledo, OH, USA |
Abstract
Abstract — A neural network is trained on simulated ellipsometric spectra to determine gradients in selenium content throughout the depth of a simulated cadmium selenide telluride (CdSe(Y)Te(1-Y)) absorber measured in the through-the-glass measurement configuration. Ten thermally evaporated CdSe(Y)Te(1-Y) alloys are deposited on TEC-12D and characterized with spectroscopic ellipsometry to develop an optical database for the alloys and substrate. This optical database can be used to train a neural network to handle ellipsometric spectra measured from real CdSe(Y)Te(1-Y) films.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.