Can chilling a column cause physical damage?

[DoryFish]

This probably sounds like a strange question, however I'm wanting only to explore a possible on-column reaction of my compound and am not worried about peak shape or resolution (at this stage). I've a compound that undergoes keto enol tautomerism and which the enol is more stable. With a Waters SunFire C18 column and an MeOH/CH3COOH mobile phase, I see (thanks to the diode array) a tailing on the keto peak which has enol UV characteristics. Eluting later is the enol peak, which has fronting with keto UV characteristics.

My theory at the moment is once the sample has separated on-column into the the keto and enol components, some tautomerism of those components then occurs. To explore this I'm altering the residence time on the column via flow rate, but would also manipulate the kinetics of the tautomerism by running at both 15 oC and 40 oC (plus the two different flow rates).

The Waters column care guide gives the temp range of 20-50 oC for "low pH in order to enhance selectivity". As I'm not worried about peak shape or resolution (at this stage), my only concern is physically damaging the column by running at < 20 oC. I can find plenty of info regards column damage at elevated temp, but nothing about the opposite. I'll gradually drop the temp so won't thermally shock it, however is shrinkage of the packing material at low temps a concern??

Thanks!

[R13]

Quick scan using Google gave some examples of low temperature HPLC separations done (also chiral separations in fully organic mobile phases at -25 and -40 C).

So temperatures below 20 C should be OK with the column. Keep in mind that at lower temperatures mobile phase viscosity increases - column pressure will be significantly higher! And of course no mixtures with water at -40 C

[DoryFish]

Fantastic!! Thanks heaps R13. I did google but am obviously incompetent .

Will certainly keep an eye on the column pressure, plus also any problems caused by fogging on the flow cell window. I live in a high humidity climate, so this could be the biggest problem.

Thanks again!

[Gerhard Kratz]

I agree with R13. As long as you do not produce ice cubes on the column it is ok.

[KDF]

Shows what I know...

I thought keto-enol tautamerism was a rapid equilibrium since it is a simple internal electronic rearrangement assisted by a general or specific acid catalyst. In fact, I though they were drawn with a delocalized electron structure. Of course, most of my kinetics training was inorganic.

Kinetically speaking you are trying to "trap" the molecules in one state or the other of the transition state energy barrier. it seems like low temperature NMR is going to tell you whether this is possible.

By the way. Pure acetic acid freezes well above 0C [18C I think]. Be careful not to get too cold.

KDF

[DoryFish]

I'm an electrochemist and thought the same about keto-enol tautomerism until started reading. We're working with a highly conjugated system with benzene rings, so apparently this stabilizes the enol over the keto form. Was never very good at organic chemistry, so please don't believe everything I write.

We ran the experiment at 10oC and amazingly it went very smoothly. Really did expect a problem with fogging on the flow cell.

With a slow flow rate and 40 oC, we saw serious fronting on the enol peak. At fast flow rate and 10 oC, almost none. Got to go through the diode array data more thoroughly, but at the moment we "think" some of the enol is converting to keto on-column.

I very much like your idea of low temp nmr. Will see how much the boss wants to prove this is what's going on. My thoughts were simply to demonstrate we couldn't couldn't improve enol peak shape (room temp) because of the tautomerism, but you're right, low temp nmr is the next step.

Thanks for sharing your ideas KDF! Very much appreciated.

[R13]

Fogging on the flow cell window might be a problem or not depending on the detector construction.

Usually UV lamp and the cell are placed inside the metal block to stabilise baseline drift/niose created by temperature changes, so cell is at somewhat elevated temperature. I think I have read somewhere, that for example Agilent 1100 UV detectors have additional constant heating for better performance.

[HW Mueller]

KDF, this tautomerism is characterized by the rearrangement of a proton, any chemistry involves electrons.

[KDF]

Yes, I am aware of the proton shift. Hence my comment of general and specific acid catalysis given the acetic acid conditions. In tautomerism, is the proton exchangeable or always the same atom? Please forgive my ignorance and/or memory loss. It has been about 30 years since I took a formal organic chemistry class. I survived mainly by understanding which way electrons want to flow.

KDF

[HW Mueller]

Since there is a acid catalysis one would think that the proton is not shifted within the same molecule.

[David Weis]

For amide H/D exchange MS work, we routinely take our columns down to 0 degrees Celsius.

Although back-pressure increases roughly 50%, we don't take a big hit in peak width with our reversed-phase separation of peptides on fast gradients.

It is worth pointing out that efficient cooling of a chromatographic column is not a trivial operation. While one can easily put a column in a refrigerated compartment, the mobile phase is typically not chilled and thus enters the column as a constant thermal load.

The equilibrium temperature reached inside the column is a competition between heat transfer into the column from the mobile phase and heat loss due to the cold air surrounding the column. The heating by the mobile phase is forced convection with a liquid while the cooling of the column is much less efficient forced air convection (if you have a fan) followed by inefficient thermal transfer through the walls of the column. A low volume pre-column heat exchanger is required to get efficient cooling. (See our app note in JASMS, Keppel et al. 2011)