IR transmission through a very thin liquid-filled cell

Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
Sign up Log in
11 replies · 2K views
hilbert2
Science Advisor
Insights Author
Messages
1,612
Reaction score
611
Hi,

does anyone know if it's possible, with a Nicolet is50 FTIR spectrometer or similar, to measure the transmittance of IR radiation through a cell filled with a liquid sample with thickness in the range of about 10 microns and accurately adjustable? The reason for doing this would be because some substances may absorb so effectively at certain wavelengths that the transmittance would be practically zero for a larger sample thickness and one couldn't calculate the absorption coefficient for the whole wavelength range from the Lambert-Beer law.

The sample would probably have to be contained in a space between some kind of IR-transparent plates, with plate distance accurately adjustable.

Thanks,
Hilbert2
 

Discussion

Chemistry news on Phys.org
Can't you dilute the sample with an inert solvent?
 
  • Like
Likes   Reactions: hilbert2
Because 10 µm is of the order of infrared wavelengths (10 µm = 1000 cm-1), your spectrum may be affected by interference fringes.
Here is a spectrum of a polyethylene film ca. 30 µ thick. At 10 µm the fringe spacing would be 3 x as great. The intensity would also depend on the refractive index mismatch at the interface.
 

Attachments

  • Like
Likes   Reactions: hilbert2 and Charles Link
Thanks, I saw those interference fringes being mentioned in some publication. The first information about this kind of measurements seems to have been produced in the 1960s: http://www.nrcresearchpress.com/doi/pdf/10.1139/v69-572 .

One approach may be to produce a liquid film with a thickness that's somewhere in the acceptable micron range, and then just measure that thickness precisely enough so that Beer-Lambert law can be used to find the absorption coefficient from the IR transmittance data.

The reason for doing these measurements is because I have to form some kind of law of how deeply blackbody-like radiation can go under a liquid surface. The radiation is produced in the gas phase above the liquid and I'm interested in something like how many millimeters of liquid are enough to absorb 90% of the total radiative energy flux - this of course depends on the blackbody temperature and the absorption spectrum of the liquid.
 
If you are not interested in modeling of a particular substance, differences between measurements of a pure liquid and a solution can be negligible - that is, they can give values good enough to construct a model.

Then, Beer-Lambert's law is not linear for concentrated solutions, but I think you should be able to find reasonably good models that could help predict real values (at least I remember being told something like that). Or - if not - perhaps you can initially choose a rather "transparent" liquid for getting some reasonable data (transparent as in "with absorption low enough to allow measurements of more thick layers").

Just thinking aloud.
 
Yeah, it's probably possible to find some relatively simple approximation that works well enough for complex hydrocarbon mixtures (like gasoline or motor oil) in most cases. A more problematic case would be a calculation for ethanol or some other alcohol diluted with water, as there the hydrogen bonding between the species will be important.
 
hilbert2 said:
A more problematic case would be a calculation for ethanol or some other alcohol diluted with water, as there the hydrogen bonding between the species will be important.

Probably, but finding spectra of pure ethanol is not that difficult. I can be missing something, won't be for the first time, but I am not convinced you really need to use thin layers for recording them.

http://sdbs.db.aist.go.jp/sdbs/cgi-bin/direct_frame_disp.cgi?sdbsno=1300

(choose IR on the left)