Denoising and Baseline Correction Prospects of Laser Induced Plasma Spectra acquired under Low Pressure Conditions: Comparison of Algorithms and Optimization Perspectives

• Published in: 2021 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT) pp. 01 - 04
• Main Authors: Sridhar, R.V.L.N., Sundram, Saibala, Sannalike, Yashas R., Upadhya, Prashanth C., Umesh, S.B., Christy Bobby, T., Malathi, S., Elumalai, S., Sriram, K.V.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 09.07.2021
• Subjects: data preprocessing; Geology; LIBS; matrix effects; planetary surface exploration; qualitative analysis; Soil; Soil measurements; Spectroscopy; Surface cleaning; Surface emitting lasers; Vacuum systems
• ISSN: 2766-2101
• DOI: 10.1109/CONECCT52877.2021.9622605

Laser induced breakdown spectroscopy (LIBS) is a versatile atomic emission spectroscopy technique for rapid multi-element analysis, and has shown potency for planetary surface chemistry studies in recent times. In spite of analytical advances, accuracy of the LIBS analytical capability is still debatable, due to unknown matrix effects i.e., random noise, continuum etc. This paper presents details on executed combinations of selected algorithms and respective parameters optimization to generate a clean spectrum from raw data sets. In this study, a feldspar soil pellet was used as a sample and LIBS spectra were acquired in a high vacuum (~1.5x10 −6 mbar) environment using a compact micro-Joule LIBS system. Elemental detection accuracy of ~ 99% is achieved with the clean LIBS spectrum generated from the proposed algorithm combinations.