Raman spectral resolution of 1.65 cm^-1 is obtained while using the 600 lines per mm grating in the spectrometer; in this condition can we compare and analyze the Raman peak shifting trends with the minimum shift of 0.5 cm^-1 which is less than the spectral resolution of the Raman data points?
Hi Mukesh,
The short answer is No. I don’t think it would be correct to monitor peak shifts that are lesser than the spectral resolution. You might have to use a higher number of gratings.
The resolution you provide is the FWHM, right? Because the actual (dimensionless) spectral reolution is defined differently, see https://en.wikipedia.org/wiki/Spectral_resolution
Anyway, with an FWHM of 1.65 cm-1 I don't see a problem about detecting a 0.5cm-1 shift.
However, in you last sentence you write something about data points. If your resolution is not the FWHM but the data point increment, then it's of course not possible, but 1.65cm-1 increments would be odd for a Raman spectrometer.
Jürgen Weippert sir, Yes, it is the data point(wavenumber) increment in a Raman spectrum, But I was amazed to see some good papers concluding the results taking the Raman shiftings (less than spectral resolution) into account.
Jürgen Weippert, could you please explain the first part of your answer a little bit more (or provide a reference) so that I can understand the FWHM aspect better? I use a 600g/mm grating with a 532 nm laser, and the exported data always has a range of 3750 cm-1 over 1600 points (there is no option for improving this, except for maybe using some interpolation). By this logic, I get a resolution of 2.3 cm-1 between data points, and therefore, any shift in peak position below 2.3 cm-1 cannot be recorded. Is this a valid argument?
The spectral resolution is not the data point increment, those are simply different quantities. The spectral resolution is, as the wikipedia article states clearly, a dimensionless quantity calculated from the mean wavelength and the peak FWHM. Since it's easier to calculate, you could also divide the wavenumber of your laser by the FWHM in cm-1. So, with 532 nm excitation, which corresponds to 18800cm-1 and an FWHM of 1.65cm-1 the resolution would be 11400.
If you had e.g. a data point increment of 0.1cm-1 which is not unrealistic and the FWHM of your peak was 1.65cm-1 you could publish a single peak shift of 0.5cm-1 without any problems; if you had two peaks at this distance, however, a separation would be somewhere between hard and impossible.
But, of course, since the "resolution" given here is not the resolution but the data point increment, you can forget about that.
Sachin Nair and Jürgen Weippert
sir, thanks a lot for your suggestions. I would like to explain the problem in addition.
There are ~7 data points in a range of 10 cm^-1, which means the difference of 1.65cm^-1 between two data points (wave-numbers), and I got a Raman spectrum for graphene, let's say G peak lies at 1575 cm^-1 when the laser-spot is located at a centre of a graphene bubble, but when I located the laser at some different location towards the edge of the graphene bubble, the G-peak blue-shifts, this is OK, but I get the shift ~0.5cm^-1. Is it reasonable to accept it and analyse the data?
So, the 0.5cm-1 come from a fit function? If you plot one G above the other, do you see a shift of the full peak or is it just the maximum? If the entire peak shifts, I think you can claim that the shift exists, but I wouldn't perform any follow-up calculations with that.
Jürgen Weippert Sir, thanks a lot for your suggestion.
yes, sir, the 0.5 cm^-1 comes from a single Lorentzian fitting. Sir, the full peak shifts to the right when stacked on top of the other. similarly, the trend follows for other locations as well. Will it be acceptable?
- Badges
- Science topic
- Similar topics
- Chemistry
- Pharmacology
- Medical Pharmacology
- Clinical Pharmacology