Chromatography Forum

I am using propylchloroformate to derivatize "theoretically pure (99%+)" amino acids to test for purity. Relative to my blank, I have four peaks (all have been background subtracted):

Peak, Retention Time (Relative Integrated area):
Peak 1, 3.77 (1.6%)
Peak 2, 4.52 (1.2%)
Peak 3, 4.87 (13.1%)
Analyte, 5.49 (84.1%)

I am using a GC/TOF. My control yielded a mass accuracy of 1.4mDa so I would assume my samples are within that range (see spectrum):



The derivatized structure looks like this with the yellow highlighted region the d(x) tags. (H is removed from carboxylic acid and replaced with C3H7 and one of the H of the nitrile is replaced with C(O)OC3H7):



As you can see from the spectrum, the three unknown peaks have some common peaks. Are these peaks from McLafferty rearrangements? Alpha cleavage of the C(O)? I have run into several amino acids that I have issues like this with. Usually on the alpha cleavage of the C(O), these are easy to spot. Anyone have any guesses on these three UK peaks? We are running these by LC/MS as well and previously with other amino acids, only one peak was detected so this would lend that these other peaks are artifacts of the derivatization process. Any ideas? Thanks!

the peak at 3.77 min results from cyclization of glutamic acid, when it looses water to form γ-lactam (5-oxopyrrolidine-2-carboxylic acid) followed by esterification of the only COOH group to give the MW 171.0895 (5-oxopyrrolidine-2-carboxylic acid propyl ester)

further, secondary amine of this lactam gets propoxycarbonyl group attached to give the product with MW 257.1263 (4.87 min)

finally, peak at 4.52 min is non-esterified N-propoxycarbonyl-5-oxopyrrolidine-2-carboxylic acid (theor MW 215.0794).

the peak at 3.77 min results from cyclization of glutamic acid, when it looses water to form γ-lactam (5-oxopyrrolidine-2-carboxylic acid) followed by esterification of the only COOH group to give the MW 171.0895 (5-oxopyrrolidine-2-carboxylic acid propyl ester)

further, secondary amine of this lactam gets propoxycarbonyl group attached to give the product with MW 257.1263 (4.87 min)

finally, peak at 4.52 min is non-esterified N-propoxycarbonyl-5-oxopyrrolidine-2-carboxylic acid (theor MW 215.0794).

Thanks Trozen so much!!! I will check these when I get back to work Monday. I had guessed the first one (3.77) but couldn't explain why. This and the other two are beyond my expertise. Can you suggest a paper/book that would give me insight into your post? Could you also explain how you know these so I can better understand myself when it comes to predicting other "artifacts" I am seeing for other amino acids. Once again...thank you!!!

the peak at 3.77 min results from cyclization of glutamic acid, when it looses water to form γ-lactam (5-oxopyrrolidine-2-carboxylic acid) followed by esterification of the only COOH group to give the MW 171.0895 (5-oxopyrrolidine-2-carboxylic acid propyl ester)

further, secondary amine of this lactam gets propoxycarbonyl group attached to give the product with MW 257.1263 (4.87 min)

finally, peak at 4.52 min is non-esterified N-propoxycarbonyl-5-oxopyrrolidine-2-carboxylic acid (theor MW 215.0794).

Trozen

I checked the accurate mass and possible fragments based on your assessment and you were spot on. Thanks once again! It was extremely helpful and will be beneficial for some other peaks. Is this a phenomena seen in amino acids with a second carboxylic and/or amino group (γ-lactam formation)? Did you learn this from experience or is there somewhere I can educate myself on these possible artifacts. I also want to thank you for the very fast reply. It made coming in Monday morning prior to an upcoming two weeks of vacation even nicer to know I could get these observations resolved before I left.

Cat5Hurricane,

sorry for coming back that late.
actually in my case it was just an experience I have got while working with chloroformate derivatives during my PhD thesis several years ago.
at that time I was reading quite a lot about this derivatization reaction which is really tricky (in terms of artifacts), but I changed the field since that time..and starting to forget all these things

anyway, I am glad to be helpful..

Trozen

Can you or anyone help me on aspartic acid as well? Very similiar in structure to the glutamic acid you helped me with but with one less CH2 group between the amine and addition COOH group:



Here are the spectrum. The analyte (aspartic acid) is the bottom trace. Accurate mass for the derivatized sample should be 303.1682. As you can see from the molecular ion peak, that is about 35 mDa off. According to the methyl stearate standard I ran prior to the run, the mass accuracy was 1.1mDa. I am a bit surprised that the molecular ion for aspartic acid is so far off as the instrument and other samples are spot on. Anyway, in the upper trace, disregard the peaks above 215 as they are seen in the blank as well. Do you have any thoughts or guesses on what these artifacts peaks are or how they are being formed? Thanks again!

I can't help that much with the spectra, but since I worked also with chloroformates derivatization I thought I might share some tips. First: this reaction is not as quantitative as sylanization, I think that if you try to change the water/basic catalizer/chloroformate ratios you might observe some changes. I worked mainly with sulfurated aminoacids (for whose this derivatization is simply the best you can get, because is very mild), but did some tests with other aminoacids too. I worked with ethyl chloroformate and never had your problem, so I think it might relate to the lower reactivity of the propyl chloroformate compared to ethyl and methyl, so that the amino group instead of preferentially reacting with the chloroformate, cyclizes in a intramolecular reaction. Second, try changing the pH of the reaction media. It is suggested often in literature that a basic pH is better for this reaction, but in my case I found the opposite. Mind, I was working with phosphate buffer solutions because I had to analyze the reaction products of an enzyme, but I found that a pH of 5 worked much better than a pH of 8 in terms of artifact formation. So I think you really have to check on a case-by-case basis. Anyway doing a couple of experiments should be pretty simple, also given that the reagents are very cheap.