Page 1 of 2
Peak shape on primesep A column
Posted: Mon Jun 02, 2014 9:53 am
by bunnahabhain
Hi everyone
I have a rather funny peak shape on my Primesep A column:
This is a primary amine run in 0.1 % TFA/H2O/Acetonitrile eluent, isocratic. I know, the peak is very small, but higher amounts result in a scaled triangle. Any idea how to improve this?
Thanks in advance
Jörg
PS.: I have seen this before analysing L-cysteine on the same column, using 0.1 formic acid in H2O/ACN, but it was much sharper compared to the peak above.
Re: Peak shape on primesep A column
Posted: Mon Jun 02, 2014 10:33 am
by Gerhard Kratz
Maybe you can speed up the cation exchange mechanism by increasing the column temperature - and increase flow rate.
Good luck.
Re: Peak shape on primesep A column
Posted: Mon Jun 02, 2014 9:55 pm
by M Farooq
Hi everyone
I have a rather funny peak shape on my Primesep A column:
This is a primary amine run in 0.1 % TFA/H2O/Acetonitrile eluent, isocratic. I know, the peak is very small, but higher amounts result in a scaled triangle. Any idea how to improve this?
Fronting occurs whenever the mobile phase component(s) strongly adheres to the stationary phase as compared to the analyte. You may try to replace TFA with some other acid, say formic acid and see if that tunes the peak shape. This phenomenon is well known in overloading studies and it has been seen in all modes of chromatography including ion exchange.
Increasing the flow rate may not alter the phenomenon, but temperature may change the competition events.
Regards.
Re: Peak shape on primesep A column
Posted: Thu Jun 05, 2014 8:45 pm
by Vlad Orlovsky
What are you analyzing and what is your detection technique? Make sure that whatever you see is your target and not impurity.
What is the X-scale and Y-scale (hard to tell from your picture).
What is sample concentration?
What is column size?
What is your mobile phase composition (amount of ACN and water)?
Once you clarify these I might be able to help you.
You can contact me by email through our website.
if you are reluctant to share information please increase TFA to 0.2% and decrease injection 5 times. See if it improves. If you are using ELSD and your amine is volatile you must use TFA. If you are not using ELSD or your amine is not volatile then you can use ammonium formate or acetate and see if peak shape improves.
Re: Peak shape on primesep A column
Posted: Tue Jun 10, 2014 2:49 pm
by bunnahabhain
What are you analyzing and what is your detection technique? Make sure that whatever you see is your target and not impurity.
This is a heterocycle with attached aminomethyl group. Sorry, I cannot give more detail here. Detection is UV 195 nm
What is the X-scale and Y-scale (hard to tell from your picture).
rather small, both 
x = 0 - 5 min, y = -10 - +20 mAU
What is sample concentration?
250 ppm
What is column size?
150x2.1 mm
What is your mobile phase composition (amount of ACN and water)?
ACN/H2O 50/50 + 0.1 % TFA
Once you clarify these I might be able to help you.
You can contact me by email through our website.
if you are reluctant to share information please increase TFA to 0.2% and decrease injection 5 times.
This might indeed help a lot. Injection volume is 100 µl.
See if it improves. If you are using ELSD and your amine is volatile you must use TFA. If you are not using ELSD or your amine is not volatile then you can use ammonium formate or acetate and see if peak shape improves.
This is a method a colleague of mine developed. She is fine with it, it is reproducible in the matrix we need it for, it passes all validation criteria, so we are going to use it as it is. However, I still have the opinion that a peak should look like a peak, therefore I was asking
. I'll keep in mind your suggestions and try it out as soon as I have to deal with a similar method
before validation is finished.
Thanks a lot for valuable comments!
Jörg
Re: Peak shape on primesep A column
Posted: Tue Jun 10, 2014 3:01 pm
by Vlad Orlovsky
I think that your main issue is injecting huge volume into narrow bore column (2.1 mm). Your retention is relatively short and high volume injection contributes to the issue. You can try to inject less and also reduce amount of ACN and TFA twice. This will move you peak allowing injection "disseminate" I also not a big fan of your 195 nm, may be running it at higher wavelength (250 nm) can help your base line issue. I assume that it is visible at higher wavelength.
Re: Peak shape on primesep A column
Posted: Wed Jun 11, 2014 9:12 am
by bunnahabhain
The analyte is not visible at higher wavelengths. at 200 nm the S/N improves slightly, but that makes not much difference. This analyte begs for LC/MS, but our MS is overloaded at the moment, so we have to go with what we have in UV. Reducing injection amount is not an option, the UV response is just too small.
Re: Peak shape on primesep A column
Posted: Wed Jun 11, 2014 9:18 am
by Gerhard Kratz
Maybe you have a micro flow cell available for your UV detector. That will increase mass sensitivity.
Re: Peak shape on primesep A column
Posted: Wed Jun 11, 2014 10:18 am
by Hollow
what is the sample solvent injected?
Same or weaker than the mobile phase? if possible try to get weaker than mobile phase.
with 100µl you actually fill about 25% (!) of your column volume with the injection, therefore be as weak as you can get.
for short I would give the reduced incetion a try: maybe with improved peak shape you will get higher peaks, balancing out the effect of the smaller peak area (?).
retention range looks quite ok to me with k of ca. 4 and 8.
with new method development, maybe try a gradient: smaller/higher peaks, possible to load the sample in as weak conditions as possible (eg. 10-20%ACN, sample and mobile phase) locking your analytes to the column, before starting to elute.
Re: Peak shape on primesep A column
Posted: Wed Jun 11, 2014 11:09 am
by bunnahabhain
Maybe you have a micro flow cell available for your UV detector. That will increase mass sensitivity.
I don't think so:
http://www.chem.agilent.com/Library/Sup ... faq111.pdf mentions decreased sensitivity with smaller flow cells. Inferior resolution is not caused by the flow cell in this case, I guess.
Jörg
Re: Peak shape on primesep A column
Posted: Wed Jun 11, 2014 1:48 pm
by danko
What are you talking about guys - TFA and 195nm detection wavelength (?) Doesn’t it ring a bell?
No wonder it’s not possible to reduce the injection volume.
Wavelength must go up – at least to 210 – 215 nm. Or the TFA must go!
Then smaller injection will be feasible.
Best Regards
Re: Peak shape on primesep A column
Posted: Wed Jun 11, 2014 2:45 pm
by Vlad Orlovsky
you can replace TFA with sulfuric or phosphoric acid. Problem is that method is validated as far as I can tell
Re: Peak shape on primesep A column
Posted: Wed Jun 11, 2014 3:14 pm
by bunnahabhain
you can replace TFA with sulfuric or phosphoric acid. Problem is that method is validated as far as I can tell
Yes, validation is finished. But still I'm happy to be able to learn for the future
danko: As the method is isocratic, and the buffer is pre-mixed, TFA's self absorbance does not impact the measurement so strongly. But you're right, it is not beautiful.
Re: Peak shape on primesep A column
Posted: Wed Jun 11, 2014 6:45 pm
by danko
quote "danko: As the method is isocratic, and the buffer is pre-mixed, TFA's self absorbance does not impact the measurement so strongly."
The above does not make sense! Of course the mixture absorbs energy as much as it would have if it wasn't premixed.
Besides it's a perfect example of a misinterpretation of the concept of validation. People expect too litle of a method because it's not woth a dime and guess what it passes. But when it comes to the real world it's worthless.
Apart from that I like your attitude of wishing to learn.
Best regards
Re: Peak shape on primesep A column
Posted: Thu Jun 12, 2014 7:20 am
by bunnahabhain
The above does not make sense! Of course the mixture absorbs energy as much as it would have if it wasn't premixed.
I agree, it does absorb the same amount of energy. What I meant is that it does not show that much the wavey baseline you see with pump-mixed eluents.
Besides it's a perfect example of a misinterpretation of the concept of validation. People expect too litle of a method because it's not woth a dime and guess what it passes. But when it comes to the real world it's worthless.
No, it's an example of different interpretation what validation can mean. It is not intended to go to the "real world" (in the meaning of transfer to other labs, different matrices etc.). It is developped for a specific use and works for that use, this is all we need. I agree, however, that the method is not beautiful and has no chance to be of much use other than it's specific purpose. I asked my question in the first place was because I don't really like this approach, too. So we are not too far away from each other, regarding our point of view on methods in general
Jörg