Chromatography Forum

Dear members of this forum,

Recently I started to investigate the area with the temperature an I looked it is increased with high temperatures (until 60ºC), I looked a very good performance of the peak.
I read an article about the influence of temperature with the peaks and mentioned the reduction of the time retention and the shape of the peak but nothing about the influence of the area. I try to know the influence of the proportion constant of Beer Law in diferent books but I did not find anything.
I hope to find anyone who give me experiences about the infuence of the temperature with the area of the peaks and what happened into the column and the detector.
Thanks in advance for your help,

Diego Delmonte

In principle, I would not expect that the peak area would change. However, there are secondary effects that depend on the temperature. The most important one is that the ionization of the compound can change, if it is an ionizable compound like an acid or a base, and you are working close to the pK of the compound. If you can exclude this first effect, there is a rather weak effect of the solvent environment on the spectrum, but this effect is really weak. Another thing that is happening is that the flow rate is a bit higher, since you heat up the solvent and expand it. This can change the peak area as well, but once again, this is a very weak effect. Of course, the last thing that can happen is that your analyte decomposes during the chromatographic run.

How much of a change do you see?

I'll add a footnote to Uwe's reply:
all of the above effects can be exacerbated if you are not working at an absorbance maximum for your analyte.

Also, at higher temperature more analyte molecules will be in higher vibratinal levels of the electronic ground state, so the absorption max will shift to longer wave length (lower energy shift to the upper electronic state). The area will go down. The Boltzmann distribution will predict this wave shift, you need to know where the vibrational levels are (from IR, Raman) very involved calcs., havn´t done this since student days. Since you need to calibrate your run it is actually a mute problem.

OK, you made the case for the absorption to shift to lower wavelength, but I didn't see the explanation for why the molar absorbtivity would also decrease. Please elaborate. Acutally, I don't know what fundamental property affects molar absorbtivity. (it is not often that I get to quiz a real spectroscopist on these fundamental issues)

PS (made SB w/ Sp. from new deer. Yum!)

Hans,
I agree with Bill. I am not familiar with this shift in spectrum with temperature, and I would like to learn more about it. How much of a shift is expected? Is it only a shift in the spectrum, but the shape remains the same? Etc..

One doesn´t see much on this, as one usually makes the measurements at room temp. There was this vague memory, so I checked my books and found something only in the old Calvert and Pitts (?, don´t have it here). Not much details were given so I am guessing here:
If you remember Jablonsky(?) diagrams, the absorption is usually shown from the lowest vibrational level of the lower electronic level to various vibrational levels of the upper electronic level (this gives the broad wavelength distribution). The max is the transition which is the most likely (involved quantum statistics...). Now the Boltzman distribution depends on, among other things, the Kelvin temp and the spacing of the vibrational levels in the ground state. For most organic molecules I would guess that a 30° difference would considerably populate higher vibrational levels in this ground state. Now the transitions to the different upper electronic level´s vibrational states still follow the same rules, so the shape of the curve should be similar, but the energy is lower (since most now come from higher vibrational states). If your detector was set at the max wave length (at room temp) then, since it moved you will be at a lower absorption. This would require that if you changed your setting to the new max your absorption would remain the same.
Shure would like to know whether this interpretation (explanation) is correct. Should have checked my statistical thermodynamics book, maybe there is an example in it, ....can still do that.

I think this is a very academic discussion and I like it. This could generate some interesting seminar exercises for physical chemistry at university.
I would not expect to see a big influence for the spectra, however: An UV absorption of e.g. 200 nm is equivalent to 50,000 cm-1 wavenumbers. If you assume a low-lying vibrational skeletal mode (200 cm-1, this is also the approximate value for kT at room temperature in Boltzmann's equation) to be higher higher populated, the then 49,800 cm-1 correspond to 200.8 nm. If you include the denser rotational energy levels, the shift still becomes smaller.
I am not sure on the influence of working in solution: This makes lifetimes very short and broadens the transitions.

I would not expect increasing temperature to significantly affect peak area, as I can't think of any reasonable mechanism for this to occur for the majority of compounds. What I have seen is a significant change in the resolution of closely eluting peaks when the temperature is changed. My first thought would be that there is now an impurity co-eluting with your analyte and increasing the area of the peak. The first time I ran a column at an elevated temperature I was somewhat surprised that I had to change my gradient as much as I had to to maintain resolution of my critical pair. I don't remember a significant change in area of the peaks.

KHW,
ok, I should have refreshed my memory a bit better by reading further in Calvert and Pitts. Even if a vibrational level of 2000+cm-1 was invoked, the shift wouldn´t be that great (~10nm). But if I understood some additional reading, one should not expect much difference in the population of the higher vibrational level, anyway, on a ~30° temp. change. One would say here in Germany that I "made the flies nervous".
Anyway, this was really academic as the change of column temperature does not correspond to an equal change in detector cell temp.

I found the discussion stimulating, none the less. But we never did get a discussion of what affects the molar absorbtivity.

PS my PChem professor was known to put the flies to sleep with his lectures. I like the "nervous fly" variant......

...Thought someone else would venture an answer here, but above I assumed that the transition probabilities (selecion rules, etc.) would not change if the transition occurred from a higher vibrational state (of the ground electronic state). That would mean that the absorption curve would shift only (absorption at a givn wavelength would not change) , and since the max wavelength shifted the absorption shown by the detector would be less. (Or: If the detector was changed to the higher temp max the absorption wouldn´t change). True?
Having burned my Fingers, I am not going to suggest a temperature change were this shift would be significant.

It's been many years since I sat through a physical chemistry lecture, but if I remember correctly the temperature effects on absorption are much more pronounced in the gas phase. The interactions with the solvent tend to turn absorbances from a relatively sharp band into a very broad band, thus for gas phase IR you needed much higher resolution that for solution IR. Given reasonable variations in temperature, I would expect to see very little change in wavelength maximum or absorbance.