![]() |
ИСТИНА |
Войти в систему Регистрация |
ИСТИНА ФИЦ ПХФ и МХ РАН |
||
Iterative approximation with exponentially modified gaussian (EMG) function is widely used to describe chromatographic peaks1. EMG has four parameters: peak height (h), position of unmodified gaussian (μ), standard deviation (σ), and relaxation time (τ)2. Peak asymmetry is most often described by the ratio of the right and the left half-width at 10% of peak height. The equations to estimate initial parameters of EMG from experimental data date back to the 1980s3. Nevertheless, modern papers often omit recommendations for setting initial values for approximation of chromatographic peaks. While the choice of initial values is not so important for EMG, it can be crucial for iterative approximation of the peak using more sophisticated functions with more parameters1. That is why we revised and improved some of the previously suggested approaches3. For example, we used interpolation with splines instead of approximation with equations, which allowed us to drastically increase the range of valid values obtained from experimental data without losing accuracy. Moreover, we explored some aspects of estimating peak apex coordinates and half-widths. We also compared the accuracy of estimation of initial parameters using half-widths at various heights (from 10% to 50%). References 1. Komsta Ł. et al. Chemometrics in Chromatography. CRC Press, 2018. 527 p. 2. Kalambet Y. et al. Reconstruction of chromatographic peaks using the exponentially modified Gaussian function // J. Chemometrics. 2011. Vol. 25, № 7. P. 352–356. 3. Foley J.P., Dorsey J.G. Equations for calculation of chromatographic figures of merit for ideal and skewed peaks // Anal. Chem. 1983. Vol. 55, № 4. P. 730–737. The research is supported by Russian Science Foundation (project No. 22-73-10053), https://rscf.ru/project/22-73-10053/
№ | Имя | Описание | Имя файла | Размер | Добавлен |
---|---|---|---|---|---|
1. | текст тезисов в сборнике тезисов конференции | materials-opt.pdf | 558,0 КБ | 24 декабря 2024 [khrisanfovmike] |