|Raman Test Report - Raman Spectrometer with Back-illuminated TE-Cooled CCD (512 x 64 Pixels), He-Ne-Laser with 35mW and 12.5mm Focus Raman Probe|
| Request for information |
|1. Description of the Measuring System|
The Raman Spectrometer examined in detail was configured for a chemical company and is to be used there for monitoring polymerization reactions. Through a glass window, the probe has to take Raman spectra in the reactor out of the reaction. Therefore, a big focal length of 12.5 mm was specified for the probe.
The entire system consists of a computer, spectrometer, laser with power adapter and probe. It has a completely modular design.
The He-Ne-laser used has an optical power output of 35mW. The coupling module for optical fibers is equipped with an FC/PC connector and allows a fairly comfortable and stable setting of the coupling’s maximum intensity into the probe.
In a longtime study of the laser a total fluctuation < 5% under lab conditions and at a period of ca. 40 hours was measured. The measurement taken was integral by coupling the laser signal via an optical fiber (constructed in the same way as the probe fiber) to an integrating sphere and the detection signal was taken with another optical fiber from the integrating sphere’s exit to a cooled CCD spectrometer for the VIS range.
As a probe, the RamanProbe was used the schematic setup of which is depicted in the illustration opposite. The probe features an internal filtering and comes with focal lengths of 1, 2, 5, 10 and 12.5 mm. The spectrometer used is a cooled CCD spectrometer with 512 x 64 pixels and a spectral range of 636nm – 811nm, corresponding to a Raman shift of 63 - 3470 wave numbers. The coupling in the spectrometer is done with a 200 m single fiber and reproduces the signal at a height of only ca. 8 of the 64 pixels (8 x 25 m=200 m).
At room conditions the detector temperature can be kept stable at -15░C. Under the selected measure-ment conditions (integration time 1s, 16 averagings), the noise is < 2 counts at the spectrometer’s measurement dynamics of 16 bit (see diagram opposite). At measured Raman peaks of up to 1000 counts, the signal-to-noise ratio of real signals is up to 500 : 1 at measuring times of ca. 16 seconds.
2. Measurement Setup for the Studies
The following test measurements were taken with the setup described above (laser – probe – spectrometer).
The probes were put in an opaque receptacle. Through its removable top the probe was lead via the samples. The probe Tip was located ca. 5mm above the liquid samples and ca. 12mm above the solid samples. The measurement was not done through the receptacle side.
During initial test measurements through the side of cylindric sample receptacles of ca. 25mm diameter, a reduction of the Raman signal up to the factor 3 was found. The integration time of the system was set to 1s; its maxium integration time is 167s. In the spetra illustrated in the appendix, 16 subsequent spectra were averaged (not coadded). They were measured and stored respectively as raw data and as energy spectra.
All spectra were stored with file headers where the measurement conditions are listed.
3. Measurements on Liquid Samples
Various substances were measured. The spectra of all measured samples can be found in the appendix. The measurements showed a high reproducibility and, over a broad range of the distance from the surface, a strong measurement signal that is nearly equal.
In the following, the results are explained as an example from the toluene spectrum. A good signal-to-noise ratio at only 1s measuring time and 16 averaged spectra is visible.
The resolution measured nearly corresponds to the above-mentioned evaluation and is ca. 20nm for the peak at 1000 wavenumbers as visible in the diagram above-right.
4. Measurements on Solid Samples
Among liquid samples, various solids were analyzed although the system, as mentioned before, is unsuitable for scattering samples and solids due to the high focal length.
When positioning the samples, it was noticed that the distance of the surface at the probe tip could vary in the millisecond range. However, a clear Raman signal was measured. Typically, for the recorded spectra a distance of approx. 10 – 12mm was set in adjusting the Raman signal to the maximum.
5. Possible System Configuration Variants
Along with the realized version, the following system options are conceivable.
- Supply with other focal lengths
- Shorter focal length should increase the signal,,
- Measuring scattering samples with a shorter focal length is probably easier
- Supply with a process probe -> Signal should be less about the factor ca. 3 (experience with excitation at 785nm)
- Spectrometer options with 1024 elements improves the optical resolution about the factor 2.
- Cross section converter in the spectrometer (e.g. 7 x 200 m) for improved degree of utilization of the detector surface, yet only useful when using respective fiber bundles or fiber cross sections for the external probe fiber.
- Supply of the system with smaller laser models (shorter module, facilitated integration)
- Supply of the system in a 19“ rack for process applications (Engineering required for laser maintenance accessibility only possible with smaller modules/ lower laser power)
- Supply as complete system with WinSpec Proc process software for on-line applications in the production plant
Last change 11/27/2007 11:09 AM
Copyright (c) getAMO 2022