Acetaldehyde split

[lvan]

I analyze carbonyl compounds on a Waters 2695 HPLC per method to11a. Acetaldehyde, which comes out at 11.492 minutes is split. Has anyone else had this happen?

[tom jupille]

My first thought was keto-enol tautomerization, but that is too rapid a process (and the equilibrium is to lopsided) in acetaldehyde. Second thought was a column inlet flow anomaly (partially-plugged frit or headspace), but that would affect all the peaks in the separation, not just the acetaldehyde. That leaves the third possibility: you have an additional compound in there that you didn't expect.

[lvan]

Thank you, Tom,
Could this be an isomer of acetaldehyde? We have been running for the last 20 years with this additional peak. It appears even when using different standards and using different columns.(lot numbers are different). Not only does the continuing calibration standard come out this way, but the calibration curve (different vendor's standard), the cartridge lot blanks and samples look the same way. All run on 2 instruments!!
Thank you for your input!!

[eg@aaclab]

You are right....The DNPHs give 2 peaks for most carbonyls except Formaldehyde..the better the separation the more chances you'll see the 2 peaks..if you do LCMS you'll see the 2 peaks are everywhere even coeluting with other peaks...there are several good papers on this...
google it....
"Since unsymmetrical aldehydes and ketones are known to form isomeric 2,4-dinitrophenylhydrazones (syn/ anti-isomers), the influence of isomerization on the practicability and accuracy of the 2,4-DNPH-method using 2,4-dinitrophenylhydrazine-coated solid sorbent samplers has been studied with three ketones (methyl ethyl ...
Syn/anti isomerization of 2,4-dinitrophenylhydrazones in the ... - NCBI

[lvan]

Thank you, EG. I appreciate your response.

[mattmullaney]

Hi Ivan,

I'm with EG above, I've run TO-11A quite a bit myself. The ketones are all symmetrical while the aldehydes are not. What you are observing is E/Z (or anti/syn) isomerism (acid-catalyzed) where the Z form is the minor peak. You can see this behavior in the standard material as well if the pH of the standard solution is overly acidic.

Polar-embedded phases reveal the isomerization most readily (such as Supelco's Ascentis RP-amide, Agilent's Bonus-RP, or Thermo's Acclaim Polar Advantage II). Are you running the method on an ordinary C18 phase?

I read mostly the work of Shigehisa Uchiyama (try ScienceDirect) about this, though it seems that the research into this area remains very active.