Two-dimensional correlation analysis of Raman optical activity - Basic rules and data treatment

详细信息	查看全文 | 推荐本文 |

• 作者：Tomá ; &scaron ; Pazderka ; Vladimí ; r Kopecký ; Jr. ; ^{kopecky@karlov.mff.cuni.cz}
• 关键词：2DCoS ; two-dimensional correlation spectroscopy ; PCA ; principal component analysis ; ROA ; Raman optical activity ; VCD ; vibrational circular dichroism ; VOA ; vibrational optical activity
• 刊名：Vibrational Spectroscopy
• 出版年：2012
• 期刊代码：113_09242031
• 类别：ms
• 出版时间：May, 2012
• 卷：60
• 期：Complete
• 页码：193-199
• 文件大小：1445 K

摘要

Two-dimensional correlation spectroscopy (2DCoS) is a powerful method which provides information about the sequential order of spectral changes in dependence on external perturbation and has very strong deconvoluting abilities. The applications in vibrational optical activity (VOA), represented by Raman optical activity (ROA) and vibrational circular dichroism, are complicated by couplets of bands of opposite signs as well as by low signal to noise ratio, background variations and strong sensitivity to baseline distortions. Here we present the comparison of simulations of wavenumber, intensity and bandwidth changes in VOA couplets and of classical bands analyzed by 2DCoS. This showed that Noda's rules are directly applicable to asynchronous part of 2DCoS but must be reversed for the synchronous part of 2DCoS. Thus, some of the spectral changes exhibit similar 2DCoS patterns which can be deduced from single band analogies. Nevertheless, other analogies may be misguided because several spectral events in VOA show similar patterns as very distant cases in classical bands. We also developed new data treatment of ROA spectral series based on principal component analysis (PCA). The core of the treatment lies in baseline corrections in PCA subspectra and subsequent usage of significant corrected subspectra for ROA data reconstruction. This removes spectral noise, background variations as well as unwanted distortions of the ROA baseline. Moreover, the proposed data treatment for 2DCoS is generally applicable to any noisy spectroscopic data sensitive to background variations.