I don't know anything about raman spectra, but to convert your plot to wave number versus raw data, you only need to change the horizontal axis labels. Let k denote wave number and L denote wavelength. Then, k=2*pi/L. Just convert every value of L on the horizontal axis to a corresponding value of k.

This, of course, will not provide raw data corresponding to equally spaced values of k, which may be what you want. For that, you will need to interpolate your data.

Dear Courtney Chng,

your suggested precedure will be the way to go.

Just keep in mind, that depending on the purpose, in spectroscopic belongings the 'spectroscopic' wavenumber is usually defined as

\tilde{\nu} = 1/\lambda

in units of [1/cm].

https://en.wikipedia.org/wiki/Wavenumber

Best regards.

Dear Courtney Chng , it is necessary first of al to digitize the graph/ It is possible to use the programm GRAPH DIGITIZER/ After it will be possible to calculate \tilde{\nu} = 1/\lambda in units of [1/cm].

After you can use for the new graph any programm, for example ORIGIN/

Best wishes

Courtney:

First, that's a rather odd looking spectrograph. I don't think I've ever seen such a linear slope in baseline like that. However, I'll ignore that for the moment. Also, your spectrum looks featureless, except for that narrow line at about 632nm, also rather odd. In Raman spectroscopy, the sample is (usually) illuminated with a laser. It is even often brought to a focus point with a lens to maximize intensity on the surface or within the sample. Light from the laser scatters both elastically (without wavelength shift) and in-elastically (with wavelength shift). This in-elastic scattering is the Raman effect.

On the detection end, the spectrometer is normally modified by adding a very narrow 'notch' filter to reject the very bright light scattered at the original laser wavelength. What remains are the photons that were in-elastically scattered at wavelengths other than the original laser wavelength.

Just how much these in-elastically scattered photons are shifted in frequency depends on the vibrations within the molecule. The higher the frequency of the vibration, the larger the shift. In essence, a Raman spectrum is a vibrational spectrum of the molecule. It is very similar to, but not identical with, a standard IR or FT-IR spectrum. The difference is, it is 'observed' at the wavelengths in the vicinity of the laser wavelength. So, if your laser is in the visible spectrum, you can get a vibrational spectrum if an aqueous specimen. Something impossible to do with a standard IR or FT-IR spectrometer because water is opaque to IR.

In the Raman effect there are two sets of spectral 'lines' generated. The most intense lines (but still very very dim) are those that lie on the red (or longer wavelength or smaller wavenumber) side of the laser line. These are called the 'Stokes' lines. There are another mirror image set of lines on the blue (or shorter wavelength) side of the laser line. These lines are called 'anti-Stokes' lines and are typically very much dimmer than the Stokes lines. Hence one normally focuses on the Stokes lines.

As for data processing, one starts by converting the laser wavelength to wavenumbers. I use a simple trick: wavenumbers (cm-1) = 1e7 / wavelength (nm). Say your laser is at 532nm. Converting that to wavenumbers: laser wavenumber = 1e7/532 = 18797 (cm-1). Now using that same formula convert all the other wavelength data over to wavenumbers. In a final step, subtract all the wavenumbers of the spectrum data from the laser wavelength wavenumber. All the data that falls on the positive side of zero are the Stokes line data. All the data that falls on the negative side of zero are the anti-Stokes line data. One normally drops the anti-Stokes data, but not always. Normally, the wavenumber data are displayed on a 0 to 4000 cm-1 scale, similar to a standard IR spectrum.

Now you have a vibrational spectrum (similar to an IR spectrum) where the peaks are in terms of wavenumber shift from the excitation (the laser). Hence one often speaks of the "Raman shift" of a vibrational line.

So Courtney, because your spectrum looks so odd to me, could you tell us how you made the measurement? I just wonder about that odd sloped baseline and the relatively featureless spectrum, which normally would not occur in a conventional Raman spectrometer.

Best regards,

Pat

Dear Patrick Bisson

Thank you for the very informative and detailed answer. I am very new to SERS and raman in general, so your reply was very helpful.

I realised my set up may have some problems, for example, the missing notch filter, which I am trying to get now.

For your information, I was trying to get raman spectra of an aqueous green tea sample on an Au substrate, with a 633nm excitation laser source. My goal is to obtain the SERS enhancement factor, but I think it will be awhile before I can achieve that!

I have attached my setup thus far.

Courtney:

OK, that's great! SERS is a good technique for enhancing the Raman effect by as much as 106. Even so, the signal is very very low in comparison to the back-scattered laser light. True, your beam splitter will reduce the laser intensity, but it will also reduce your Raman signal. If your splitter is 50/50, then you'll loose 50% of your Raman signal too. In addition, the relative ratio of laser line intensity to Raman line intensity will not be effected. That's bad.

Given your setup diagram, I'd place the notch filter between the splitter and the lens. This is not an ideal configuration. Another approach, which is more optically efficient, is to replace the cube-type beam splitter with an 'edge filter", or perhaps a dichroic beam splitter/combiner (AKA: dichroic mirror) where the cutoff is just a few nm longer than your laser wavelength. Turn it 450 to the laser beam.

The filter will reflect the laser onto the surface just like the splitter, but allow the slightly longer Stokes lines to pass on to your spectrometer. This approach has the advantage of directing nearly 100% of the laser energy on to the sample, and still allowing nearly 100% of the Stokes lines to pass on to your spectrometer.

The trick is to get that filter edge as close to the laser wavelength as possible. For example, with a 633nm laser line, you might pick a 650nm long pass dichroic mirror (available from Thorlabs). This filter would block the laser line almost completely. However, you'd loose the Stokes lines that fall in the 633nm - 650nm region. That's about a 500 cm-1 Raman shift. Normally, shifts as low as that would correspond to molecular bending and libration modes. One normally looks at proper vibrational stretching modes which lay at higher wavenumbers. But it depends on the experiment.

On the other hand, you can buy a specialty filter from Semrock (who specializes in very steep, sharp filters). They happen to have several with an edge specifically designed for a 632.8nm laser line.

Don't get me wrong, I'm not a salesman. There are a number of different suppliers of filters out there. I'm just making some suggestions.

Courtney, I hope this all helps!

Pat

Dear Courtney Chng,

yes, as the colleagues say, convert the wavelength to wavenumbers according to the formulas they give. Just decide what you mean under the wavenumber 1/lambda or k. The difference is the factor of 2pi. This is enough, but unfortunatly, after the conversion you get your spectrum defined on a nonlinear grid. Sometimes you may want to have it uniform.

Suppose you have your spectrum recorded on N wavelenght steps from lambda(1) to lambda(N). You can define a new uniform wavenumber axis having N equal steps from 1/lambda(N) to 1/lambda(1). Then you can use a nice matlub function "interp1"(see Matlab Help) to redifine your spectrum. Just do not forget to renormalize the spectral components. To fulfill the energy conservation you have to divide the result at each step "i" with the wavenumber(i)^2, which is the derivative d(wavenumber)/d(lambda) .

Regards,

AT

Just an important remark: the line you see on your spectrum is the Rayleigh band of the excitation wavelength not any Raman signal!

Since you work with an aqueous solution, you should see a line at 820 nm (Raman band of the O-H vibration), so you are not looking in the right wavelength range! If you can detect some organics in your sample, a C-H vibration should be seen around 775 nm and a C=O vibration around 704 nm.

If you cannot measure the spectrum after 780 nm, then change the excitation wavelength. Note that most PMs have low sensitivity at wavelengths larger than 750 nm; you have to worry about this.

Taking a 532 nm excitation laser would lead to bands at 658, 629, and 581 nm for the vibrations mentioned above.

The 633nm line shown in Courtney's spectrum is indeed the Rayleigh scattering (elastic) of the laser. The strong, broad line that would appear in water, with a laser line at 633nm would be at a 3400 cm-1 offset. That translates out to a 778nm wavelength (Stokes line). Courtney's spectrum does extend out to >800nm so she should be detecting it under normal conditions.

However, Courtney did not use a 'notch' filter or an 'edge' filter to exclude the laser line. Hence, no doubt to avoid saturating the spectrometer, the integration time was adjusted to be very short. This precludes the detection of the many orders of magnitude dimmer Stokes lines.

Note, the 'transparent' window for water in Raman is from 0 cm-1 (well, the edge of the filter) out to about 3900 cm-1. It's not truly transparent, but it's just that there are no strong Stokes lines there to interfere with anything else in the water from showing up.

Unless or until Courtney adds either a notch filter or an edge filter to exclude the strong laser line at 633nm, she won't be able to detect any Stokes lines, even with SERS.

@Patrick; I think that the Raman band of water should be seen at wavelength >800 nm.

633 nm correspond to 15798 cm-1; deducting 3400 (or 3600 as I did) leads to 12398 or 12198 cm-1, translating to 807 or 820 nm.

@Jean-Claude,, yes you are right. My mistake (I hate it when I do that!). Even so, it looks like Courtney's spectrometer reaches out to 880nm. That would be a Raman shift of more than 4000 cm-1, enough to encompass the entire (conventional) vibrational spectrum. So there's no danger of missing any Stokes lines. Given the diagram of the setup, it is only missing a filter of some sort to knock out the laser line.

Hello eveyone, thank you so much thus far for your valuable inuts.

I have added a filter as per Patrick's suggestion, and the rayleigh scattering of the laser is out.

Previously, there was also a downward sloping baseline, which does not appear anymore. I can't explain this, it just suddenly happened ._.

This time I tried illuminating a sample of ethanol (95%) methanol (5%) with the 633 laser, and obtained the attached graph "newest.jpg". I am very confused as to how I am supposed to deduce the peaks. (Fyi, Integration time is at 6s and averaging is 4). I tried to smooth the same graph out with a smoothing factor of 6 as seen in the attached graph "smoothing is 6.jpg". I don't know what else could be wrong with the experiment, but I hope to troubleshoot as much as possible and obtain a decent raman spectra before moving on to obtaining the SERS one.

Thank you everyone once again.

Courtney:

OK, making progress I see. Raman will never be a 'quick' measurement. The signal is always so low that long integration times are pretty much the norm. Rather than us smoothing at this point, I'd take greater averaging factors. This increases the signal-to-noise ratio (S/N) by the square root of the averaging factor. So, if you average by 9, S/N increases by 3. If you average by 100, S/N increases by 10.

It looks like your signal counts are pretty good. Try lowering your integration time a bit. Say to about 5 or even 4 seconds. But increase the averaging count to 10 or even 20. Time consuming, I know. But that's the name of the game.

Let me suggest stop using the alcohol as a prototype sample until you get your setup working. Instead, change back to DI water (e.g., 18M Ohm water). As Jean-Claude pointed out earlier, there's a huge fat, whopping peak at about 3400 cm-1 offset (about 806nm in your setup). Water's always my 'go-to' sample for testing Raman.

Also, could you post another sketch of your modified setup? I just want to make sure I understand what you did. Did you use a notch filter, edge filter ... what?

Pat

Hi Patrick

Attached is an image of my modified setup.

I'm using a highpass filter that only allows 700nm & higher to pass through. Parameters are also changed to:

Integration Time - 6s

Averaging - 20

Smoothing - 6

After switching out the alcohol sample for DI water as you suggested, I realized something is wrong with my experiment as the measurement taken without any sample, and with a sample, are the same (maybe except for the graph shifting either up or down). The same plot shape appears when I tried to take measurements with a pure Si wafer as well. [see DI Water.JPG and DI Water with substrate.JPG]

The spectrometer is definitely detecting though as the distinct peak of the 633 laser appears on the plot when I remove the filter.

Does this mean that the raman signal is too weak? Is the noise masking the raman signal? I am not entirely sure what I can do to get the respective peaks to appear.

The measurement were taken with the lights off and laser-blocking curtains drawn. The environment is not entire pitch black, but is relatively dark.

Also, I've read reports where they incorporated CCD. Is this truly necessary if the spectrometer has an in-built CCD?