Page 1 of 1
Thermodynamic Pitfall(s)
Posted: Sun Dec 25, 2005 1:35 am
by Ashraf_Khan
Is there any theromdynamic pitfall(s) in gas chromatography ?
Posted: Sun Dec 25, 2005 5:02 am
by tom jupille
There is lots of thermodynamics, but what do you mean by "pitfalls"?
Thermodynamic Pitfall(s) in Chromatography
Posted: Tue Dec 27, 2005 5:52 pm
by Ashraf_Khan
Thanks.
The word pitfall(s) speaks for itself. By thermodynamic pitfall(s) I meant whether there is any ambiguity when thermodynamic parameters, such as reference state, equilibrium constant, partition coefficient et al., are applied to chromatographic processes.
Posted: Tue Dec 27, 2005 7:40 pm
by tom jupille
Sorry, if the word "pitfalls" really spoke for itself, I wouldn't have inquired

.
One can understand and predict
retention on the basis of system thermodynamics. This is extremely valuable on a qualitative (approximate) level, but becomes problematic if you want to make accurate predictions (there are typically multiple interacting processes at work and there is a slight theoretical complication in that there is no
true equilibrium because one of the phases is moving).
Peak broadening , on the other hand, is controlled largely by system kinetics.
Thermodynamic Pitfalls
Posted: Wed Dec 28, 2005 2:54 am
by Ashraf_Khan
Thanks for your input (Sorry, the word pitfalls was ambigious itself). I understand your point regarding the retention values.
Are the 'chromatographic' partition coefficient and the 'thermodynamic' partition coefficient same thing ? The former is based on the solute concentration, whereas the latter based on the solute activity (which takes into account the stoichiometry, intermolecular interactions et al.).
Posted: Wed Dec 28, 2005 8:27 pm
by tom jupille
Are the 'chromatographic' partition coefficient and the 'thermodynamic' partition coefficient same thing ?
Strictly speaking, the answer is "no". My recollection from thermo class (many years ago) is that activity and concentration approach identity at infinite dilution. That said, the analyte concentration is usually sufficiently low that the "infinite dilution" assumption works well enough for practical purposes (i.e., the distribution isotherms are reasonably linear).
We're approaching the limits of my knowledge here, so if any one else wants to chime in and/or correct me, please jump in!