Identification of Unknown Nanofabrication Chemicals Using Raman Spectroscopy and Deep Learning
- Resource Type
- Periodical
- Authors
- Theobald, N.; Ledvina, D.; Kukula, K.; Maines, S.; Hasz, K.; Raschke, M.; Crawford, J.; Jessing, J.; Li, Y.
- Source
- IEEE Sensors Journal IEEE Sensors J. Sensors Journal, IEEE. 23(7):7910-7916 Apr, 2023
- Subject
- Signal Processing and Analysis
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Robotics and Control Systems
Containers
Chemicals
Glass
Raman scattering
Solvents
Resists
Nanofabrication
Convolutional neural nets
deep learning
machine learning
nanofabrication
Raman spectroscopy
- Language
- ISSN
- 1530-437X
1558-1748
2379-9153
Raman spectroscopy is a common identification and analysis technique used in research and manufacturing industries. This study investigates the use of Raman spectroscopy and deep learning techniques for identifying various nanofabrication chemicals. Four solvents and SU-8 developer were identified inside common chemical storage and distribution containers. The containers attenuated the spectra and contributed varying amounts of background fluorescence, making manual identification difficult. Two varieties of SU-8 photoresist were differentiated inside amber glass jars, and cured samples of three ratios of polydimethylsiloxane (PDMS) were differentiated using Raman microscopy. The neural network accurately identified the nanofabrication chemicals 100% of the time, without additional preprocessing. This investigation demonstrates the use of Raman spectroscopy and neural networks for the identification of nanofabrication chemicals and makes recommendations for use in other challenging identification applications.