1) GC-columns

Problem:

Among analytical chemists you may often find the following statement:

"Non-polar GC-columns separate organic compounds according to their boiling point."

Note:
(i) In this context, non-polar means apolar.
(ii) The boiling point of organic compounds is typically proportional to the logarithm of their saturated (liquid) vapor pressure at a given temperature.

Therefore the statement is equivalent to:

"Apolar GC-columns (e.g. DB5) separate organic compounds according to (= in the order of) their saturated (liquid) vapor pressure at temperature x."

1a) Is this first statement correct?

Similarly, you will also find the statement:

"Polar columns (e.g. Carbowax) separate organic compounds according to their polarity."

1b) What do you think about this second statement?

Answer:

1a) On an apolar column, compounds are retained only due to their van der Waals interactions (in an apolar stationary phase, there are only van der Waals interactions between the phase molecules and the analyte). Clearly, a correlation between logarithmic retention and the vapor pressure can be expected ONLY for those compounds whose vapor pressures also solely depend on van der Waals interaction. That is, for compounds which in their pure liquid states only have van der Waals interactions with one another. This is true for apolar and monopolar compounds. For bipolar compounds (i.e., compounds with both H-donor and H-acceptor properties) the above correlation will break: the vapor pressures of these compounds are also lowerd by H-bonds (in addition to the ubiquitous van der Waals interactions) in their pure liquid phase. However, these interactions will not affect their retention on an apolar stationary phase (see also Chapter IV.3.4 in the script).

1b) This statement is too simplistic. Although there are no commercial GC columns with H-bond donor properties, there are still two different kinds of polar interactions possible on GC columns (those that are described by the A and S descriptor of the analyte). In addition, all stationary phases will also retain compounds due to their van der Waals interactions (remember: these always operate). However, so called "polar" columns have such a high cohesive energy that the required cavity energy (i.e., the energy required to make a cavity in the phase for the analyte) cancels much of the van der Waals energy betweeen the analyte and the phase. Therefore, retention won't show a strong dependence on molecular size.