Hello, I'm still learning the ways of FTIR and when I was trying to do real-time FTIR using an ATR device (ZnSe crystal), my spectra came out "delayed" (for lack of a better word) or shifted in position to a lower wavenumbers.
The peak positions are simply off (see picture, bottom spectrum) compared to when scanning it individually (see picture, top spectrum). I could also could use advice on how to reduce the annoying noise. Thank you!
Specs: 32 scans @ 4 cm-1 resolution for 60 seconds using SmartITR, using MCT/A detector cooled with Liquid N2.
Sample: Methacrylate functionalized fatty acid with photoinitiator, a drop on the crystal cured using an UV LED spot curing system.
Dear Harris:
1. there is no surprise that the peaks in ATR spectrum are shifted to lower vawenumbers as compared with common spectra collected in transmissional mode - BECAUSE by contrast to common spectra - the ATR spectra depend on vawenumber:
(the penetration depth(dp) in ATR is increasing at lower frequencies. And Dp in ATR is like an equivalent of "t" - thickness in common IR spectra Thus it is natural that band positions n ATR are shifted to lower frequencies. The same caused the change in band-shape. For example the symmetric (in transmission) peak slghtly broadens from the side of lower vawenumbers.
But the most important thing is the ratio of refractive indices of your substance and the internal reflection element (IRE). ATR works well when the ratio n2/n1 , where n2 is the refractive index of IRE and N1 - that of your sample. The higher n2 - the better the ATR spectrum.
Among the optical materials used for IREs the highest n2 =4.0 has Ge; the lowest n2=1.8 ZnSe. And it is the VERY PROBLEM of employing as IRE ZnSe !!!
If you use ZnSe (and you can not change the "ATR rules" - all the peaks in your ATR spectrum will be:
a) distorted (it concerns band shapes, especially for strong peaks)
b) that is in particular why - some of your strongest peaks (and you should know that at maximum of strong IR peak the refractive index is much higher AND much closer to the refractive index of ZnSe - the reason why all the maxima of strong peaks are like "cut off"
The very very strong peaks may have (at maximum) the refractive index so close to that of ZnSe - that the strong band could disappear at all.
Indeed, imagine the situation when n1=n2 - there will be NO internal reflection at all
and the IR beam will travel through the IRE-crystal - directly through the sample -- to nowhere.
My strong advice - in order to have no such a problems in future - you should replace ZnSe by KRS-5 (n2=2.4) or diamond (n2 is also 2.4)
All the best, Leonid
You do not have noise. That is water within the gas phase. Flash the equipment with dry nitrogen before and during the measurement.
And if everything works fine the spectra are not significantly shifted.
As I understand the blue one is onto the prism and measured before the illumination and the red one is one (the first?) of the real-time measurements of the SAME sample. Hence you have a problem with your soft-, hardware. I guess the software/settings and not the hardware:
1. The water spectra are “transmission” spectra (your noise)
2. The first signal of water around 3700 cm-1 is not shifted
3. The second water signal around 1500 cm-1 is shifted
As far as I know the service of the distributer is quite good.
Which vibration do you want to analyse 1630-1680 or 880-995. You may be interested in some of our publications (Raman curing experiments and multiple ZnSe spectra of differently modified acrylic esters.
Article Introduction to analyzing the solidification of multifunctio...
Article Synthesis and testing of ZnO nanoparticles for photo-initiat...
Article Method to analyse energy and intensity dependent photo-curin...
I agree with Michael Schmitt. It is rather a "problem" of software than really shift of the peak position. Both spectra just are schown in different scales of wavelength.
- Badges
- Science topic
- Similar topics
- Correction