Suppose you have an infrared absorption spectrum with a huge peak at some wavelength or wavenumber. Then there’s another much smaller peak at a wavenumber twice as big (wavelength half as long). Would you typically assume the smaller peak represents real physical absorption by the sample? Or would you suspect an instrumental artifact?
Suppose more particularly you were comparing several spectra of the same substance, published by different researchers. And suppose the size of the smaller peak varies quite a lot from researcher to researcher, while the huge peak is nicely consistent.
A spectroscope of the old type that uses prisms wouldn’t do this, but if the spectroscopy is Fourier transform spectroscopy, and the detector has a nonlinearity of only 0.1%, wouldn’t you see an alias peak 3 orders of magnitude smaller in actual absorbed energy? For that matter, wouldn’t a grating (in less-ancient spectroscopes) create its own aliases (for other reasons unrelated to sensor linearity)? How do spectroscopists deal with that effect (if I imagine it correctly)?
I don’t use FTIR much, but can’t peaks shift due to stresses in the sample. If the 2nd peak is exactly 1/2 the wavelength, it is possible that there’s a probably with the machine. On the other hand I could be talking compete bollocks!
In my own graduiate work I did see absorptions at twice the wavenumber and three times and four times, etc. A lot of vibrational levels were, in fact, present.
But in the real world, since potentials aren’t perfectly parabolic, you have anharmonicities, and so the higher absorptions aren’t exactly twice the wavenumber. In fact, the anharmonicities are typically linear themselves, over short ranges. So it was easy, in my case, to distinguish between instrumental and physical effects.
You get the same harmonic effects in grating spectrometers as in FTIR devices. It’s true that you don’t see them in prismatic instruments, but you’re trading the relatively simplemath in gratings or FTIR for the complex index variation of your prism. Plus prism spectrometers are a lot harder to find these days.
The time-honored solution is to put a blocking filter in to remove higher orders.
So long as the transitions are not forbidden, you can often see both harmonic and combination transitions. Sometimes only even or odd harmonics are allowed, depending on symmetry. I’m not sure what sort of samples you’re looking at (could you let us know?), but with most organic molecules you can see both fundamental and harmonic absorptions for C-H, C-C, C=C, etc. stretches (I think–you may want to pick up an orgo book if it’s important to be sure of this.) The harmonics (and lots of funky combination transitions) are often in the “thumbprint” region of the absorption spectrum.
Eh, what Daftbugger said. Impurities and solvents (both of which can be further messed with by temp and pressure) can do funny things.
I don’t know much about detectors. Listening to CalMeacham is probably wise .
Note that I only have one undergrad pchem classes and one undergrad spectroscopy seminar, so I may be wrong and others may know a lot more about all this.
.