Pls follow the work from CRF of Sandier national laboratory.

Do you need remote measurement of concentration? Which type of cells do you investigate? Biological cells? Which geometry has your measurement set-up? Which gases and in which concentration range do you need to control?

Typically CARS is multi-beam, multi-source technique, i.e. CARS need relative complicated experimental set-up. Measurement results are sensitive to external factors, e.g. temperature.

Most gases have strong absorption in IR (methane at 1.5,  2, 3 mkm ). Have you consider to use for your problem IR absorption spectroscopy, photo-acoustic spectroscopy or DIAL techique?

Thanks for the answers (and questions).

We have nothing biological at all. We have in mind to measure the concentration profile of methane as function of distance (maybe ~100µm) from a wall. All is in context of combustion research/ engines. We are just starting to think about it. I also thought about IR absorption but in this case we would still need the spatial resolution (which will probably pretty bad). Photo-acoustic spectroscopy might also pretty hard to do in an engine.

I already found some old papers from the Sandia national laboratory by R. Lucht. Seems like they measured the N2 concentration already with 25µm resolution back in 1987, even without imaging.

Conventional Raman scattering is probably not sensitive enough to do anything. That's why I am wondering about the sensitivity of CARS...

You are welcome!

If you need 2D concentration profile, then absorption spectroscopy is principally  limited by focussing spot diameter of laser beam (down to several mkm). Sensitivity of absorption methods is easy to estimate, and response is linear an independent from temperature.

In some tasks LIBS is well suited for concentration profiling (not only of solid state but also gases).

Thanks. I will look into this.

I found a publication of the Sandia Labs where they actually did 2D CARS imaging of N2 in a flame with rotational resolution. So it seems to be possible (although maybe not that easy).

In case of application of imaging to IR absorption, it might really be an issue with the availability of good CCD cameras for mid-IR. 

Do you can not use scanning? Than LIBS isn't adequate for you task.

For absorption  imaging of methane can be used near IR - 1.5 mkm  with InGaAs camera.

Only absorption cross section of methane at 1.5 mkm is weak (relative to 3 mkm). On other hand there are many diode lasers at 1.5 mkm and this wavelength is eye-safe.

If we scan, it is certainly an option.

Ideally, we would like to get the complete distribution in one laser shot. I also found a camera for the mid-IR region from 1-5µm which would enable using the nu3 fundamental at about 3.3µm. There are p.e. quantum-cascade laser available that operate in this range. Otherwise, one can use nonlinear optics to make pulsed IR light in that region.

Thanks for the answers. It seems like we have quite some options. 

You are welcome!