TY - JOUR

T1 - On the potential and limitations of multivariate curve resolution in Mӧssbauer spectroscopic studies

AU - Debus, Bruno

AU - Panchuk, Vitaly

AU - Gusev, Boris

AU - Savinov, Sergey

AU - Popkov, Vadim

AU - Legin, Andrey

AU - Semenov, Valentin

AU - Kirsanov, Dmitry

PY - 2020/3/15

Y1 - 2020/3/15

N2 - Traditional processing of Mӧssbauer spectroscopy measurements assumes a decomposition of the spectra into separate multiplets corresponding to particular non-equivalent states of the resonance atom. When the number of spectra is large (e.g. in kinetic, corrosion and phase transition studies), this procedure becomes time-consuming. Moreover, traditional processing assumes some hypotheses on the number of non-equivalent states and initial multiplet parameters. The results of the processing strongly depend on these hypotheses and may be quite subjective. In an attempt to circumvent this issue, we studied the potential of Multivariate curve resolution (MCR) to unravel mixed multiplets spectra into their individual contributions. The application of MCR to Mӧssbauer studies was found to be quite challenging due to 1) long acquisition times limiting the number of available samples, 2) presence of critical spectral overlaps and 3) occasional deviations from the ideal bilinear assumption. In this report, we show how these limitations can be circumvented under certain conditions.

AB - Traditional processing of Mӧssbauer spectroscopy measurements assumes a decomposition of the spectra into separate multiplets corresponding to particular non-equivalent states of the resonance atom. When the number of spectra is large (e.g. in kinetic, corrosion and phase transition studies), this procedure becomes time-consuming. Moreover, traditional processing assumes some hypotheses on the number of non-equivalent states and initial multiplet parameters. The results of the processing strongly depend on these hypotheses and may be quite subjective. In an attempt to circumvent this issue, we studied the potential of Multivariate curve resolution (MCR) to unravel mixed multiplets spectra into their individual contributions. The application of MCR to Mӧssbauer studies was found to be quite challenging due to 1) long acquisition times limiting the number of available samples, 2) presence of critical spectral overlaps and 3) occasional deviations from the ideal bilinear assumption. In this report, we show how these limitations can be circumvented under certain conditions.

KW - Chemometrics

KW - Data processing

KW - Multivariate curve resolution

KW - Mӧssbauer spectroscopy

KW - ALTERNATING LEAST-SQUARES

KW - MIXTURE ANALYSIS

KW - Mossbauer spectroscopy

KW - EVOLVING FACTOR-ANALYSIS

KW - CONSTRAINTS

KW - AMBIGUITY

KW - RANK

KW - CHROMATOGRAPHY

UR - http://www.scopus.com/inward/record.url?scp=85078510266&partnerID=8YFLogxK

U2 - 10.1016/j.chemolab.2020.103941

DO - 10.1016/j.chemolab.2020.103941

M3 - Article

AN - SCOPUS:85078510266

VL - 198

JO - Chemometrics and Intelligent Laboratory Systems

JF - Chemometrics and Intelligent Laboratory Systems

SN - 0169-7439

M1 - 103941

ER -