-
- admin
- Site Admin
-
- Posts: 647
- Joined: Wed Aug 11, 2004 12:19 am
By Uwe Neue on Thursday, May 27, 2004 - 03:57 pm:
Rob, I agree with the responses that talk about the sample solvent as the problem (Amouse, Greg)It is also common that the least retained compound is more sensitive to this phenomenon than the more retained compound(s). The reason for the fact that some columns are less sensitive than others has something to do with the details of the end-fitting design. The endfitting that creates more mixing will mix the sample better with the starting conditions of your gradient, and the occurance of this problem is less likely. On the other hand, the endfitting with the larger mixing also causes lower plate counts for the very early peaks, so there is a compromise.
If you want to know more about this effect, I recommend my book "HPLC Columns" on page 360 in the troubleshooting section.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Friday, May 28, 2004 - 12:06 am:
Rob,
well, with an amine you could have a sample-mobile phase pH incompatibility as well. One can imagine that the TFA acid interaction with the stat. phase is quite different for two different columns and thus influence the equilibrium (or nonequilibrium) between the amine and ammonium differently. Of course, the mixing mentioned by Uwe can also be effective here.
Anon May 27 7:59,
reversing the column often, also, "repairs" a void, so no proof here.
-------------------------------------------------------------------------------------------------------
By syx_gf on Friday, May 28, 2004 - 12:56 am:
We have experienced with the same phenomena while using USP method for cefadroxil oral suspension in QC Lab. The method is using a mixture of pH 5.0 phosphate buffer and acetonitrile (96:4) as mobile phase. The lab used LiChrosphere RP18 column and got the split peak problem after several injection. I recommend using Atlantis dC18 column because of high percentage of aqueous solution in mobile phase but the phenomena is still occur. Then I try the method in my lab using Atlantis dC18 column and we have no problem. Until now I do not have any answer for this mystery.
-------------------------------------------------------------------------------------------------------
By Anonymous on Sunday, May 30, 2004 - 02:31 pm:
This is just a general statement and is not intended to answer the original question, but sample solution viscosity can effect peak shape due to mixing.
-------------------------------------------------------------------------------------------------------
By Anonymous on Sunday, June 6, 2004 - 12:40 am:
I am pretty sure that the buffer strength of your eluent is not sufficent to turn all the 2° amine into the corresponding TFA salt. Increase TFA concentration up to 0.1% (we usually work with this concentration and have a success chromatographing basic compounds on C18 columns) or switch to strong buffer (i.e. phosphoric if your detection technique allows non-volatile mobile phase components) or ion pairing reagent. The very nice opportunity is to switch to different column, for example I recommend in your case column with enbedded ion pairing reagent (Primesep 100, 200). This column allows your to alalyse compounds of very different polarity in short in most cases isocratic run, because the column applies two different separation mechanisms simultaneously: ion exchange and ordinary RP, so you can selectively affect your amine RT by changing acid modifier concentration and affect non-polar compounds RT's by changing organic solvent content.
-------------------------------------------------------------------------------------------------------
By Anonymous on Friday, June 11, 2004 - 12:13 am:
Try to inject less volume.
Start with 5 µl and check if peak splitting occurs. If okay, try to increase in 5 µl steps.
Rob, I agree with the responses that talk about the sample solvent as the problem (Amouse, Greg)It is also common that the least retained compound is more sensitive to this phenomenon than the more retained compound(s). The reason for the fact that some columns are less sensitive than others has something to do with the details of the end-fitting design. The endfitting that creates more mixing will mix the sample better with the starting conditions of your gradient, and the occurance of this problem is less likely. On the other hand, the endfitting with the larger mixing also causes lower plate counts for the very early peaks, so there is a compromise.
If you want to know more about this effect, I recommend my book "HPLC Columns" on page 360 in the troubleshooting section.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Friday, May 28, 2004 - 12:06 am:
Rob,
well, with an amine you could have a sample-mobile phase pH incompatibility as well. One can imagine that the TFA acid interaction with the stat. phase is quite different for two different columns and thus influence the equilibrium (or nonequilibrium) between the amine and ammonium differently. Of course, the mixing mentioned by Uwe can also be effective here.
Anon May 27 7:59,
reversing the column often, also, "repairs" a void, so no proof here.
-------------------------------------------------------------------------------------------------------
By syx_gf on Friday, May 28, 2004 - 12:56 am:
We have experienced with the same phenomena while using USP method for cefadroxil oral suspension in QC Lab. The method is using a mixture of pH 5.0 phosphate buffer and acetonitrile (96:4) as mobile phase. The lab used LiChrosphere RP18 column and got the split peak problem after several injection. I recommend using Atlantis dC18 column because of high percentage of aqueous solution in mobile phase but the phenomena is still occur. Then I try the method in my lab using Atlantis dC18 column and we have no problem. Until now I do not have any answer for this mystery.
-------------------------------------------------------------------------------------------------------
By Anonymous on Sunday, May 30, 2004 - 02:31 pm:
This is just a general statement and is not intended to answer the original question, but sample solution viscosity can effect peak shape due to mixing.
-------------------------------------------------------------------------------------------------------
By Anonymous on Sunday, June 6, 2004 - 12:40 am:
I am pretty sure that the buffer strength of your eluent is not sufficent to turn all the 2° amine into the corresponding TFA salt. Increase TFA concentration up to 0.1% (we usually work with this concentration and have a success chromatographing basic compounds on C18 columns) or switch to strong buffer (i.e. phosphoric if your detection technique allows non-volatile mobile phase components) or ion pairing reagent. The very nice opportunity is to switch to different column, for example I recommend in your case column with enbedded ion pairing reagent (Primesep 100, 200). This column allows your to alalyse compounds of very different polarity in short in most cases isocratic run, because the column applies two different separation mechanisms simultaneously: ion exchange and ordinary RP, so you can selectively affect your amine RT by changing acid modifier concentration and affect non-polar compounds RT's by changing organic solvent content.
-------------------------------------------------------------------------------------------------------
By Anonymous on Friday, June 11, 2004 - 12:13 am:
Try to inject less volume.
Start with 5 µl and check if peak splitting occurs. If okay, try to increase in 5 µl steps.
