Modifications required for faster Inline detection

[Lincoln20XX]

Massive oversimplification incoming from a very non-expert:

From what I can tell from what I've read, it seems like the biggest single determinant of detection speed in a given Raman system is the power of the laser used to do the excitation. The detector and other components play a part, of course. But it seems to me that the biggest data sheet difference I can find on systems that are set up for online detection and have short detector times seems to be the applied mW of laser power. 450mW to 500mW seems to be a pretty common number.

So... if I wanted to modify a 532nm OpenRAMAN to be an effective inline/online detector, would it be as simple as "apply a bigger laser to the problem?" Obviously, one would also need a flow cuvette, a way to make the liquid flow through the cuvette, etc.

Am I missing some other hardware factor that complicates this?

[andy]

Luc can give a much more informed opinion but,

The issue with high powered lasers is you are pumping a lot of light energy into a tiny focused dot so you end up instantly evaporating your sample or destroying it.

There is actually a list of things you can change though:

From this book: McCreery R. L. (2000). Raman spectroscopy for chemical analysis. John Wiley & Sons.

• (Signal magnitude for a sample... how well a material sends back a raman signal)
• (measurement time)
• (laser power in photons per second / area of laser beam)
• Sample depth (for backscattering type systems, may also depend on focus)
• Transmission of optics [slit size and other losses]
• Angle of collection (i.e. how much of the emitted Raman signal is collected)
• Quantum efficiency of detector (i.e. how much of the light is captured into a digital signal)
• Dark noise of detector (which can be reduced through cooling)

What a lot of practitioners recommend is actually trying to get a better signal-to-noise by using a deeply cooled spectrometer. Keep in mind that even a very weak laser will emit some raman-scatted photons, its just a matter of can you capture those photons and record them. The problem we are dealing with is that there's a lot of background noise caused by heat. And that background noise often overshadows the handful of raman photons. So you either have to reduce the background noise by deeply cooling the sensor, or blast a lot more raman photons, either by a stronger laser, better optical design, or bigger sample.

A practical issue you have though is for the flow cuvettes they are usually square where as the OpenRaman liquid cuvette is round. this link talks about why Luc chose that but it was because test tubes are cheap and disposable whereas square quartz cuvettes are like $40 each. But practically speaking this means you would need to design a new cuvette holder for the flow cuvette.

Another thing to keep in mind is what you are trying to measure, how much raman signal that it produces, and are you trying to measure something in high enough concentration to be readable. At least for isopropyl my tested limit was like 6-60ppm for 100mw at 5seconds. So if you want less ppm then that you may have to look into SERS which is a chemical technique for amplifying the signal. It is not practical for inline I don't think.

[Luc]

Something that I can add also is to take into account the fluorescence of the supporting medium.

At the maximum exposure time of the camera, 32 seconds, I reach the fluorescence of water at ~50 mW so it makes no sense to use more power unless I plan to take shorter exposure time. That's if you want to get the maximum number of photons but you can use less and still recognize spectrum. Pure acetone is already recognizable at 100 Hz 5 mW laser power!

Of course, bigger laser = more expensive to keep the same FWHM for emission.

Concerning the flow cuvette, it's in my todo list

[Lincoln20XX]

Thank you very much for the informational replies gentlemen.

I had a feeling I might be in over my head on this one, and while I'm not exactly happy that I was right, I'm glad you helped me realize that before I started throwing money in a poorly-chosen direction.