Raman data acquisition

The bottleneck in utilizing Raman shifted rapidly from data acquisition to data interpretation. Visual differentiation works well when polymorph spectra are dramatically different or when reference samples are available for comparison, but is poorly suited for automation, for spectrally similar polymorphs, or when the form was previously unknown [231]. Spectral match techniques, such as are used in spectral libraries, help with automation, but can have trouble when the reference library is too small. Easily automated clustering techniques, such as hierarchical cluster analysis (HCA) or PCA, group similar spectra and provide information on the degree of similarity within each group [223,230]. The techniques operate best on large data sets. As an alternative, researchers at Pfizer tested several different analysis of variance (ANOVA) techniques, along with descriptive statistics, to identify different polymorphs from measurements of Raman... 

The thickness of pharmaceutical tablet coatings was predicted using target factor analysis (TFA) applied to Raman spectra collected with a 532-mn laser, where the samples were photobleached in a controlled manner before spectra were acquired. The authors acknowledge numerous issues that limit the direct applicability of this approach to process control. These include potential damage or alteration of the samples from photobleaching, laser wavelength selection, and data acquisition time. However, most of the issues raised relate to the hardware selected for a particular implementation and do not diminish the demonstration [286]. 

Fig. 5.6. A block diagram of an optical coherence tomography/Raman spectroscopy system C, circulator RSOD, rapid scanning optical delay BP, 785 bandpass BSO, beam shaping optics DM1, dichroic mirror at 990 nm DM2, dichroic mirror at 800-950 nm LP, long pass at 808 nm GP, galvanometer pair BD, balanced detector BPF, electronic band-pass filter AI-AO DAQ, analog input-output data acquisition (reprinted with permission from [34]. Copyright 2008 Optical Society of America)...

Irrespective of the method chosen, meaningful data can only be obtained if the appropriate level of signal to noise (S/N) is reached in the spectrum of each analyte. This has been achieved for Raman measurements through short data acquisition times (<1 s) and application of mathematical approaches such as If-harmonic means clustering (KHMC), factor analysis [57] and principal component analysis (PCA) [58] to the data set. Ultimately the sample response to the excitation energy determines the speed that a measurement can be made. 

The initial studies described above were conducted on Raman microspectrometers on extracted human teeth. In order to translate this research from the laboratory bench to the dental chair side, the next step it is to develop dedicated systems for clinical use. A key element to this development is the use of fibre-optic probes to allow measurements in vivo. We recently reported a study in which optical fibres were used for PRS measurements [48], Although not yet fully realized into a dental probe, this study demonstrated the design and feasibility of acquiring parallel- and cross-polarized Raman spectra via a bifurcated optical fibre whose distal terminal has the two fibres aligned vertically for simultaneously collecting spectra from the two polarization channels on a 2D CCD array. Simultaneous data acquisition will allow for more rapid measurement times in vivo. 

Figure 3. Schematic of turbulent combustor geometry and optical data acquisition system for vibrational Raman-scattering temperature measurements using SAS intensity ratios. Also shown are sketches of the expected Raman contours viewed by each of the photomultiplier detectors, the temperature calibration curve, and several expected pdf s of temperature at different flame radial positions. The actual SAS temperature calibration curve was calculated theoretically to within a constant factor. This constant, which accounted for the optical and electronic system sensitivities, was determined experimentally by means of SAS measurements made on a premixed laminar flame of known temperature. Measurements of Ne concentration were made also with this apparatus, based on the integrated Stokes vibrational Q-branch intensities. These signals were related to gas densities by calibration against ambient air signals.

Also described in Ref. k is a new optical layout for LV data acquisition which permits a significant increase in the overlap between the Raman and LV probe test volumes. The worth of the various correlations of density and temperature with velocity is critically dependent upon the accuracy of this overlap at all flame measurement positions. Thus, one must either lock the Raman and LV probes together in a precise but movable fashion -a rather difficult procedure for the precision required for bench scale" laboratory flames - or else translate the flame. 

The process of measuring the difference between the two Raman parent spectra (right and left) is shown on the flow chart of the ROA data acquisition program in Figure 13. This program is written in Array Basic supported in Spectra Calc software. First, three spectral memory banks are created for the current, Raman parent and ROA spectra. Then several coadded spectra at four different quarter-wave plate positions are taken to complete one QWP cycle. The... 

Figure 13. Diagram illustrating the data acquisition method for the computer-controlled measurement of Raman and ROA spectra at Syracuse University.

Fig. 12. SERS spectra of (a) an SAM film of PySH and (b) an alternating bilayer self-assembly film of PySH and RuPc, and (c) a Raman spectrum of RuPc in bulk. The laser power at the sample position was 1.0 mW for (a) and (b), and the data acquisition time was 90 s. The laser power at the sample position was 4.0 mW for (c) and the data acquisition time was 150 s. (Reproduced with permission from Ref. [47]. Copyright (2004) Elsevier.)...

SERS spectra in Ag hydrosols were recorded using a Jobin-Yvon HG2S monochromator equipped with a cooled RCA-C31034A photomultiplier and a data acquisition facility. To reduce the thermal effects due to the laser light, a defocused beam with low power (20 mW) was used. Raman data were obtained with exciting lines supplied by Ar"- and Kr"- lasers (406.7, 413.1, 457.9, 488.0, 514.5, 520.8, 568.2, 647.1,676.4 nm) or by He-Ne laser (632.8 nm). AU spectra were corrected to account for monochromator and photomultiplier efficiency. Power density measurements were performed with a power meter instrument (model 362 Scientech, Boulder, CO, USA) giving 5% accuracy in the 300-1,000 run spectral range. 

Now, the big difference between multichannel and multiplex techniques arises because a single detector in FT-Raman is collecting all the modulated wavelengths at once. Consider a single data acquisition in an FT experiment, with the interferometer at a fixed position. The detector collects light from... 

For FT-Raman spectrometers, an equivalent one-point calibration is more reliable because interferometers are less prone to mechanical errors. Nearly all interferometer designs include a well-defined reference wavelength (often a He-Ne laser at 632.8 nm), which is used to control data acquisition. In addition, observed FT frequencies are calculated from a large number of individual measurements, so minor mechanical jitter and random timing errors are averaged out. Provided the laser and reference frequencies are known accurately, an observed FT-Raman frequency is quite accurate, and the one-point calibration is usually adequate. 

Returning to the dispersive case, it is far more reliable to use many calibration lines than to use only one. Ideally, a large number of accurately known frequencies would be dispersed across the spectrum, then observed under the same conditions as the sample. Assuming the optical and data acquisition conditions are precisely reproduced for the standard and the sample, the sample frequencies may be accurately calculated from the standard spectrum. As noted earlier, Raman shifts may then be determined from the equally accurately known laser frequency. 

Due to the strong evaporation of some melts, fast recording of the spectra was necessary. In this case, the Raman spectrometer was interfaced with an IBM PS/2 30 microcomputer. The software controlled the spectrometer, the data acquisition, and the extraction of useful information from spectra, such as subtraction, smoothing, de-convolution, etc. In order to present the data on the screen in real time, 2200 measurements per second were recorded. With one point per cm the scan rate as fast as 1000cm Vmin could be achieved with a reasonable signal to noise ratio. 

---