Analysis and improvement of asymmetric absorbance peak models for IR spectroscopy
| dc.contributor.author | Harrysingh, Christopher | |
| dc.contributor.examiningcommittee | Zhang, Qiang (Biosystems Engineering) | |
| dc.contributor.examiningcommittee | Erkinbaev, Chyngyz (Biosystems Engineering) | |
| dc.contributor.supervisor | Morrison, Jason | |
| dc.date.accessioned | 2024-08-30T20:29:50Z | |
| dc.date.available | 2024-08-30T20:29:50Z | |
| dc.date.issued | 2024-08-27 | |
| dc.date.submitted | 2024-08-27T13:41:41Z | en_US |
| dc.date.submitted | 2024-08-30T20:09:54Z | en_US |
| dc.degree.discipline | Biosystems Engineering | |
| dc.degree.level | Master of Science (M.Sc.) | |
| dc.description.abstract | Access to spectral interpretation software has barriers, including prohibitive cost, poor software support, and little diversity in spectral reference libraries used to identify compounds. OpenSpecy is a free, open-source software tool created expressly to address these obstacles. Despite its success, the limitations and errors of OpenSpecy have not been well-studied. This study created the korepanovupdate R language package, to investigate the mathematical foundations of OpenSpecy, especially regarding the asymmetric lineshape models of Stancik and Brauns used to estimate individual peaks. The need for area normalization functions for Stancik and Brauns’ model was recognized and derived through trial and error in this study. This thesis extends the analysis of Korepanov and Sedlovets’ model and examines the asymmetric lineshape models of both sets of authors. This study lowers the error range for area normalization of Korepanov and Sedlovets asymmetric Gaussian-based lineshape from (-0.230%, 0%] to (- 1.47×10-11%, 3.87×10-11%). Similarly, the Lorentzian-based lineshape’s error range was lowered from (-0.0468%, 0.0164%) to (-5.08×10-12%, 4.80×10-12%). This study also provides the first area normalization of Stancik and Brauns’ asymmetric lineshape model, drastically reducing this model’s Gaussian area estimation error range from (-2.43%, 45.67%) to (-0.372%, 0.398%), and its Lorentzian area estimation error range from [0.00%, 57.84%) to (-0.806%, 0.506%). The results of this study also provide the first estimation methods for peak width (g) and asymmetry (a) parameters in Korepanov and Sedlovets’ asymmetric lineshape model. These estimates enable nonlinear least-squares optimization to accurately fit a peak with Korepanov and Sedlovets’ model. Lastly, it was found that two symmetric peaks can be fit to Korepanov and Sedlovets’ asymmetric model. | |
| dc.description.note | October 2024 | |
| dc.description.sponsorship | NA | |
| dc.identifier.uri | http://hdl.handle.net/1993/38478 | |
| dc.language.iso | eng | |
| dc.rights | open access | en_US |
| dc.subject | Infrared Spectroscopy | |
| dc.subject | Peak fitting | |
| dc.subject | Peak modelling | |
| dc.subject | Asymmetric peaks | |
| dc.subject | Pseudo-Voigt | |
| dc.title | Analysis and improvement of asymmetric absorbance peak models for IR spectroscopy | |
| dc.type | master thesis | en_US |
| local.subject.manitoba | no | |
| oaire.awardNumber | NA | |
| oaire.awardTitle | University of Manitoba Graduate Fellowship (Masters) | |
| oaire.awardURI | NA | |
| project.funder.identifier | https://doi.org/10.13039/100010318 | |
| project.funder.name | University of Manitoba |