Accurate determination of the onset wavelength (λonset) in optical spectroscopy. (May 2021)

• Record Type: Journal Article
• Authors: Wallace, Austin M.; Curiac, Christine; Delcamp, Jared H.; Fortenberry, Ryan C.
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 265(2021)
• Publication Date: 2021-05
• Subjects: Spectral lineshapes -- Computational algorithms -- Electronic spectroscopy -- Data processing; Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2021.107544
• Languages: English
• ISSNs: 0022-4073

Abstract

Highlights: A systematic, mathematical approach for attributing onset wavelength is established. Most manual attributions contain small errors. The range of errors, though, can be as great as 30 nm. Graphical abstract: Image, graphical abstract Abstract: Human attribution of absorption onset wavelength ( λ onset ) most often produces small, but non-negligible errors in the attribution of this value tied to the optical energy gap. The present work utilizes a free and freely available computer program ("0nset") developed herein to determine the inflection point and subsequent x-intercept defining λ onset . The manually attributed absorption curve λ onset wavelengths are typically in error to the red (in 27 out of 38 attributions) for the dataset utilized in the present work when a rigorous manual analysis was performed implying a systematic human error. The raw ( − 2 nm) and absolute (3 nm) errors are relatively small, but there are exceptions when the errors are even larger than 20 nm. The numerically computed results from 0nset are independent of application bias and will reduce errors moving forward by consistently generating λ onset values from the same algorithm. Additionally, it will reduce errors in circumstances where the inflection points are difficult to isolate or in regions like the UV where ∼ 5–10 nm errors are more significant energetically. Finally, 0nset only requires .csv inputs and is built for both Windows and Unix-based operating systems making uptake and usage straightforward for experimental groups.