Raman data analysis

To better understand the perks and pitfalls of Raman data analysis, an understanding of the characteristics of Raman spectra used for analysis is required. If data analyses are classified into two categories, qualitative and quantitative, an easy delineation of goals is provided. 

The period under review has seen a small, but apparently real, decrease in the annual number of publications in the field of the vibrational spectroscopy of transition metal carbonyls. Perhaps more important, and not unrelated, has been the change in perspective of the subject over the last few years. Although it continues to be widely used, the emphasis has moved from the simple method of v(CO) vibrational analysis first proposed by Cotton and Kraihanzel2 which itself is derived from an earlier model4 to more accurate analyses. One of the attractions of the Cotton-Kraihanzel model is its economy of parameters, making it appropriate if under-determination is to be avoided. Two developments have changed this situation. Firstly, the widespread availability of Raman facilities has made observable frequencies which previously were either only indirectly or uncertainly available. Not unfrequently, however, these additional Raman data have been obtained from studies on crystalline samples, a procedure which, in view of the additional spectral features which can occur with crystalline solids (vide infra), must be regarded as questionable. The second source of new information has been studies on isotopically-labelled species. 

The IR spectra of M2(CO)jo M = Mn,Re,Tc with particular emphasis on 13CO-containing species. A comparative vibrational analysis is given (including Raman data)... 

Since a larger sample volume is presumed to be probed, the use of transmission mode has led to simpler, more accurate models requiring fewer calibration samples [50]. Scientists at AstraZeneca found that with a transmission Raman approach as few as three calibration samples were required to obtain prediction errors nearly equivalent to their full model [42]. For a fixed 10-s acquisition time, the transmission system had prediction errors as much as 30% less than the WAI system, though both approaches had low errors. It is hoped that this approach in combination with advanced data analysis techniques, such as band target entropy minimization (BTEM) [51], might help improve Raman s quantitative sensitivity further. 

L. Zhang, M. Henson and S. Sekulic, Multivariate data analysis for Raman imaging of a model pharmaceutical tablet. Anal Chim. Acta, 545(2), 262-278 (2005). 

In our view, it is extremely difficult to use the Raman data to dismiss any structure of the norbomyl ion. The main reason for this is that a normal-coordinate analysis of the comer-protonated... 

Figure 3. Raman data obtained for the analysis of 75 15 10 P(M-OM-CN) ter-polymer. Key 0.2-W power at 514.5 nm, 4-cm bandpass, 2-cm step size, 10-s...

In the pmr data for the terpolymer, overlap between the CH3 absorption of the oxime ester and the backbone absorption is greater than in the copolymer pointed out in Figure k. Thus, while the agreement between the Raman and pmr data for the terpolymer is not very good, (lT-32 difference), it is completely within the experimental error of the pmr data. This large error and the fact that pmr can only distinguish two of the components of the terpolymer demonstrate that it is unsuited for compositional analysis of this system. Based on the agreement with published reactivity ratios and with the elemental analysis of the P(M-CN) copolymer, it is assumed that the Raman data are more accurate. 

Kini (2002). The Raman data are based on the works by Sum (1996), Tulk et al. (1998), Subramanian (2000), and Hester (2007). To accompany these tables, select NMR and Raman spectra are given in Figures 6.13 through 6.19. These tables and sample spectra should serve as a useful reference to those embarking on spectroscopic measurements and analysis of clathrate hydrates. 

Conventional microbiological identification of isolates from patients can normally be obtained with a total turnaround time of 48-96 h. Ibelings et al. [106] and Maquelin et al. [46] developed alternatively a Raman spectroscopic approach for the identification of clinically relevant Candida species from smears and microcolonies in peritonitis patients taking at least overnight (smears) or about 6h (microcolonies). Hereby, a prediction accuracy of 90% was obtained for Raman spectroscopy in combination with multivariate statistical data analysis. 

The feasibility of diffuse reflectance NIR, Fourier transform mid-IR and FT-Raman spectroscopy in combination with multivariate data analysis for in/ on-line compositional analysis of binary polymer blends found in household and industrial recyclates has been reported [121, 122]. In addition, a thorough chemometric analysis of the Raman spectral data was performed. 

Raman spectra in the range 1545-355 cm 1 were selected for data analysis. An average of 27 (461/17) spectra were obtained for each individual with a 3 min integration time per spectrum. Each spectrum was obtained with excitation power 300 mW and integration time equivalent to 3 min. Spectra from each volunteer were analyzed using PLS with leave-one-out cross-validation, with eight factors retained for development of the regression vector. For one subject, a mean absolute error (MAE) of 7.8% (RMSECV 0.7 mM) and an R2 of 0.83 were obtained. 

---