Towards removal of water vapour and carbon dioxide effects from mid-infrared spectra
Towards removal of water vapour and carbon dioxide effects from mid-infrared spectra
Fluctuation in the amount of water vapour and carbon dioxide in the beam path
in purged Fourier transform infrared (FTIR) systems following measurement of a background spectrum has a detrimental effect on the quality of spectra; the resulting decreased SNR has the potential to confuse machine learning model calibration such that it ignores these ‘fingerprint’ regions of the spectra which may contain important information pertaining to the analyte. Examples of analytes that can be affected by water vapour effects on spectra are proteins and lipids which have major characteristic absorbances in the region 1500 – 1800 cm−1. Postulated solutions to this problem include physically introducing dry or humid air to restore conditions to those at which the background was taken [1], Savitsky-Golay filtering and spectral subtraction. However, filtering results in loss of spectral information [2], whilst spectral subtraction using a separately measured water vapour spectrum is flawed unless conditions in which all spectra are taken match [3]. We present a simple method to correct for water vapour and carbon dioxide absorption, for both normal incidence and ATR techniques, using three spectra collected in close succession of the same sample, as shown in Figure 1(a). The first and last spectra are designated as references used to generate a difference spectrum by subtraction. The difference spectrum is subtracted from the analyte uncorrected spectrum to give an output spectrum (Figure 1(b)), the length of which is minimized by using a fitting function to optimise the difference spectrum scaling factor λ. Successful processing of spectra with absorption peaks, such as that of sphingomyelin, has been demonstrated utilising this technique. This methodology enables spectral correction without the requirement for significant additional sample collection or complex data processing.
Ahmed, Waseem
bd8069b9-786f-4dd1-b3f0-cf94d828a3b2
Osborne, Eleanor Louisa
02a010d9-0022-4886-9862-84317076adc2
Senthil Murugan, Ganapathy
a867686e-0535-46cc-ad85-c2342086b25b
21 April 2024
Ahmed, Waseem
bd8069b9-786f-4dd1-b3f0-cf94d828a3b2
Osborne, Eleanor Louisa
02a010d9-0022-4886-9862-84317076adc2
Senthil Murugan, Ganapathy
a867686e-0535-46cc-ad85-c2342086b25b
Ahmed, Waseem, Osborne, Eleanor Louisa and Senthil Murugan, Ganapathy
(2024)
Towards removal of water vapour and carbon dioxide effects from mid-infrared spectra.
PhotonIcs and Electromagnetics Research Symposium (PIERS) 2024, , Chengdu, China.
21 - 25 Apr 2024.
1 pp
.
Record type:
Conference or Workshop Item
(Paper)
Abstract
Fluctuation in the amount of water vapour and carbon dioxide in the beam path
in purged Fourier transform infrared (FTIR) systems following measurement of a background spectrum has a detrimental effect on the quality of spectra; the resulting decreased SNR has the potential to confuse machine learning model calibration such that it ignores these ‘fingerprint’ regions of the spectra which may contain important information pertaining to the analyte. Examples of analytes that can be affected by water vapour effects on spectra are proteins and lipids which have major characteristic absorbances in the region 1500 – 1800 cm−1. Postulated solutions to this problem include physically introducing dry or humid air to restore conditions to those at which the background was taken [1], Savitsky-Golay filtering and spectral subtraction. However, filtering results in loss of spectral information [2], whilst spectral subtraction using a separately measured water vapour spectrum is flawed unless conditions in which all spectra are taken match [3]. We present a simple method to correct for water vapour and carbon dioxide absorption, for both normal incidence and ATR techniques, using three spectra collected in close succession of the same sample, as shown in Figure 1(a). The first and last spectra are designated as references used to generate a difference spectrum by subtraction. The difference spectrum is subtracted from the analyte uncorrected spectrum to give an output spectrum (Figure 1(b)), the length of which is minimized by using a fitting function to optimise the difference spectrum scaling factor λ. Successful processing of spectra with absorption peaks, such as that of sphingomyelin, has been demonstrated utilising this technique. This methodology enables spectral correction without the requirement for significant additional sample collection or complex data processing.
Text
PIERS_abstract_revised_final
- Other
More information
Published date: 21 April 2024
Additional Information:
Invited
Venue - Dates:
PhotonIcs and Electromagnetics Research Symposium (PIERS) 2024, , Chengdu, China, 2024-04-21 - 2024-04-25
Identifiers
Local EPrints ID: 497682
URI: http://eprints.soton.ac.uk/id/eprint/497682
PURE UUID: d8ab1a6a-ec8a-4427-a8cf-96e4f3114afe
Catalogue record
Date deposited: 29 Jan 2025 18:24
Last modified: 30 Jan 2025 03:26
Export record
Contributors
Author:
Waseem Ahmed
Author:
Eleanor Louisa Osborne
Author:
Ganapathy Senthil Murugan
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics