-
- admin
- Site Admin
-
- Posts: 647
- Joined: Wed Aug 11, 2004 12:19 am
By AlfredL on Tuesday, June 1, 2004 - 10:00 pm:
Dear HPLC professionals:
I have had a problem with bad tailing for a while with our system, and I couldn't solve it.
Here is the description of the system: It consists of a UV detector, an autosampler, 2 pumps, one controller, a PC, and everything is controlled by the software titled "Class VP."
I have tried to reduce injection volume; replace all new fittings, and filter; replace some shorter tubing; flush with IPA; replace column (new vendors) and other tricks, but all did not work.
I have checked the specifications of the system, and they are:
Sampling loop: 50 uL.
Dead volume (ave from my experiments): 90 uL - this is high
Detector cell volume: 8 uL.
Cell path length: 10 mm.
Outlet tubing is 0.13 x 600 (OD is 1/16 in.)
Any suggestion is welcome. I am anxious to try other approaches, to move forward.
Thank you.
-------------------------------------------------------------------------------------------------------
By R C on Wednesday, June 2, 2004 - 06:43 am:
You've done lots of things so far but we need more specific information to help you. What is the sample -- if proprietary, what is the general class of compound? What is the column -- manufacturer, model, dimensions and it's chemistry? What is your mobile phase -- and how is it prepared?
Basically, we'll use this information to see how sound your method is. To give the worst case scenario, if you're analyzing a compound with multiple basic groups, on an old fashioned ODS column, using poor buffers -- it will take a lot of work to get results acceptable to you again.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Wednesday, June 2, 2004 - 07:20 am:
What do you call bad tailing?
Your column to the detector connection is 600mm long??
-------------------------------------------------------------------------------------------------------
By Henrik Vogelius on Wednesday, June 2, 2004 - 08:15 am:
Is the problem just arrived or has it alwayes been there, is it the first time your application run on this system.
Try to change the gradient a bit and also wash with HNO3 for 30 min, (remember to flush with H2O before and after HNO3, and also to remove column)
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Wednesday, June 2, 2004 - 01:05 pm:
What type of tubing are you using? Are you cutting the tubing yourself? If so, what method are you using? Often tailing can be created by tubing which is not cut squarely. If you are using stainless steel tubing then you probably are already using precut tubing and this shouldn't be an issue. But if you are using PEEK tubing, one has to be mindful of the cutting method in order to assure properly cut tubing is obtained. Upchurch makes a tubing cutter that is used as a rotary cutter which will generally make good-quality cuts. To cut tubing using this tubing cutter, one you should rotate the tubing cutter around the tube 3-4 times and then snap the tubing at the point where the tubing was scored by the tubing cutter. Generally, it's not a bad idea to inspect the end of the tubing using an inspection microscope to make sure there weren't any significant defects introduced into the tube ends, as even with this method there can sometimes be defects.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Wednesday, June 2, 2004 - 04:04 pm:
How do you know that it is not the column? What column are you using? Also, you can check the width of the peak by injecting a sample without the column. With a extra-column dead volume of 90 microL, I expect the peak width (at 4.4% of the height) to be also around 90 microL, maybe just a bit better. If it is much worse, see how the injecotr is plumbed (I assume that you are not injecting the full 50 microL of the injection loop???).
-------------------------------------------------------------------------------------------------------
By Anonymous on Friday, June 4, 2004 - 06:01 am:
How about getting some of the standard that the column manufacturer uses to test their column (ususally a test sheet accompanies a column) and trying to repeat their run... If that works, then it is most likely a method development issue.. if not then perhaps theres a problem with your column or system.
-------------------------------------------------------------------------------------------------------
By DR on Friday, June 4, 2004 - 06:46 am:
More details are needed, as others have stated. 90µL is not bad for dwell volume, especially in a Shimadzu. Modular ones typically run ~2mL for modular high pressure binary gradient systems.
-------------------------------------------------------------------------------------------------------
By AlfredL on Wednesday, June 9, 2004 - 09:21 am:
Dear all professionals:
Thank you for all your responses. I will try to go back in time, and provide some more info.
The name of the compound is Oxymetazoline HCl (it has a monograph in USP 27). The mobile phase is MeOH:H2O:NaAc:HAc = 40:46:10:4.
We usually analyze it using a Supelco SCX column (4.6 x 250mm) on an older HPLC system (using a separate detector, and integrator). We also have some tailing problems on this one, especially when the column is almost dead. The sampling loop of this old system is 200uL.
We would like to transfer the method to the newer Shimadzu, which is controlled by a PC. But we observe problems.
Supelco SCX column (4.6 x 250) gave peak tailing. T ~ 2.4-3.
Phenomenex SCX (4.6 x 250) gave high back pressure.
Waters Spherisorb SCX (4.6 x 150) gave good peak shape at the beginning, but later on peak height decreased, and bad tailing (T > 3).
Since I would like to validate the method with a SCX column, I have tried several things, but none is satisfactory. (This is not a comprehensive list).
1.Reduce sample injection: I reduced from 20 to 10, to 3-4 uL (on the old system, I would inject 20 uL as the SOP).
2.Flush column and check for aging: The Waters SCX column was new; nonetheless, I have tried to regenerate using several solvents (as recommended by mfg), but the peak height was only improved for a few runs right after the regeneration.
3.Check fittings and connections: I replaced all the old PEEK fittings. Also, replaced the filter. No change.
4.Check contamination: I flushed the system with IPA, and then switched back to mp. No improvement.
5.Check for dead volume: I ran the Shimadzu system without column; water as mp at low flow rate (0.2 ml/min), and inject Acetone to measured the dead volume. Ave. was 90 uL. The peak had bad tailing also (T =2.6-3.4).
6.Check for secondary interactions. I believed that there could be some absorption on to the column due to uncapped Si, as well as some affinity due to the metal ions (present in the silica support).
I had to search several vendors to find low-cost SCX column with gel support (instead of Si base). I found two: Hamilton and Higgins.
a.The Hamilton column (3.2 x 150 ?) gave a chromatogram with no clear baseline, so I assumed that it is because of the pore size (100A). [A similar situation could be found on pp. 240 in the HPLC book by J. Swadesh, ISBN 0849300037]. Also, this column is made to analyze metal ions chiefly.
b.The Higgins column (3.2 x 20) needs a holder, but it reduced run time significantly. Peak tailing is still 2.1-2.2 > 2.0.
I then believed that the SCX columns (especially the Higgins) might not be the problem. I had to look into the system.
7.Check for plumbing: I looked into the manual, and found that the mfg installed tubing with ID 0.13mm (very long outlet tubing indeed). Visual inspection told me that it could be 0.3 or 0.5 mm ID (but the Errata of the manual stated that it is 0.3 mm ID). I decided to order precut tubings with ID 0.127 mm to replace current outlet tube, and more if I could (from inj. Point to 6-position valve).
8. One last thing I should have mentioned is that I have run a C8 column, with the premixed test solution as sample. The chromatogram was not similar to the mfg’s, and I observed peak tailing as well.
Since I am waiting for the arrival of the tubes, I cannot predict if it will solve our problem. But I will post the outcome.
Thanks for reading this long list.
Alfred.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Wednesday, June 9, 2004 - 04:55 pm:
What was the peak width of your acetone injections without column (measured at 4.4% of the peak height)?
-------------------------------------------------------------------------------------------------------
By DR on Thursday, June 10, 2004 - 01:02 pm:
I've used SCX columns in the past - Mine always had a nasty habit of voiding at the inlet. This quickly leads to profound tailing, splitting etc. I suggest that you get a cartridge guard column and a new analytical column. Inspect the guard column regularly (or keep close tabs on its back pressure w/o the analytical column in place) to make sure you have some spare packing between the pumps and the analytical column.
good luck,
DR
-------------------------------------------------------------------------------------------------------
By AlfredL on Thursday, June 10, 2004 - 03:56 pm:
Peak width of the Acetone peak at 4.4% peak height is 0.2428 min. (ave.), or equiv. to 48.6 uL.
-------------------------------------------------------------------------------------------------------
By AlfredL on Saturday, June 12, 2004 - 08:46 am:
I have replaced the outlet tubing (it has smaller ID and is shorter), but the tailing is not improved. Checking the system w/o column by injecting Acetone (mp = water), the tailing is still observed (but less than before).
What should I try next?
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Saturday, June 12, 2004 - 03:50 pm:
Based on all of your comments above, I think most likely the tailing you are observing is due to secondary interactions between your analyte and contaminants on your stationary phase. Generally, the effect of "plumbing" problems should be most pronounced for early using peaks. Even if you are observing an asymmetry value of 2.6-3.4 without a column, an analyte with a decent k' should exhibit significantly lower asymmetry values whereas you are sometimes observing comparable asymmetry even when the column is present. Since you report that under some conditions (at least initially) you observe respectable asymmetry, this would seem to suggest that the problem with tailing is related either to your injection solvent, your column or some sort of contaminant which strongly interacts with your analyte. Of these choices, I can't rule out the possibility that your problem is sample solvent related since you haven't revealed the specifics of your sample solvent. But the fact that you have seen the problem on different columns to varying extents suggests at least part of the problem is column related. The fact that the Waters column gave good peak shape it at least initially also supports that view. If you want to unambiguously sort out whether the problem is sample related, "plumbing related" or due to a column contamination problem. Try the following:
1. Make up your sample with as little solvent as possible and minimize the ionic strength of your sample.
2. Observe the effect of increasing retention on analyte peak symmetry by progressively reducing the ionic strength of your mobile phase.
If changing your sample solvent has a major impact on peak tailing than the problem is probably related to your sample solvent. Ideally, in liquid chromatography the sample should be dissolved in an "eluent" with lower elution strength than your mobile phase. Under these conditions, analyte focusing will occur during the loading step, minimizing peak distortion caused by "plumbing problems" in your injection system.
If changing the sample solvent has no effect, then varying the retention of your analyte well I you to ascertain as to whether or not the problem is caused by extra column affects (in this case the asymmetry will improve as the retention time increases) or caused by stationary phase effects (in this case the asymmetry will either stay the same or get worse as retention increases). If you observe the asymmetry increasing as the retention time increases, this would seem to narrow the problem down to either your stationary phase itself or some kind of stationary phase contamination. One possible contamination of a SCX column which might cause these effects for your analyte is iron contamination. This might explain why you observed acceptable symmetry initially on the Waters column but later saw degradation of peak shape. To test the metal contamination hypothesis, I would recommend treating the Waters column with 0.1 molar oxalate (pH 2.5) overnight. If you see improve peak shape after this treatment you can be fairly sure the problem is metal contamination related. In this case you might want to install a chelating trap column ahead of the injection valve to eliminate metals arising from system contamination.
-------------------------------------------------------------------------------------------------------
By AllsepTech on Sunday, June 13, 2004 - 07:11 am:
You might also consider our approach to tailing problems. We have recently developed a method for oxymetazoline with symmetry very close to 1.0. There is also a comparison with two leading brands of columns. Check this link, I think this will help:
http://allsep.com/makeChr.php?chr=Chr_065
-------------------------------------------------------------------------------------------------------
By AllsepTech on Sunday, June 13, 2004 - 08:16 am:
Oxymetazoline is a very strong base and tailing is probably caused by secondary interactions with residual silanol groups. If you are using end-capped column this tailing is less pronounced at the beginning, but during the life time of the column the endcapping is slowly coming off and you will observe more tailing. Non-endcapped columns will give you bad shape even at the beginning. We are using a different approach, our stationary phases do not have endcapping instead the silanol interactions are completely masked by our stationary phase. Even with strong bases like oxymetazoline and quaternary amines you will observe a symmetrical peak:
http://allsep.com/makeCmp.php?cmp=Cmp_114
http://allsep.com/makeCmp.php?cmp=Cmp_110
-------------------------------------------------------------------------------------------------------
By Anonymous on Monday, June 14, 2004 - 12:48 pm:
I am trying to analize betaine and carnitine but have a very bad peak shape using XTeraa column (it slghtly better on Polaris and Zorbax colomns, but still not acceptable by my internal requirements). Both compounds are internal quaternary amines. I tried different ration AC-Water and MeOH-water with buffers, but still have a triangle peak. How can I improve peak shape for this compounds?
-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, June 17, 2004 - 07:59 am:
You still got tailing w/o column, and therefore I still think you should clean the system with 5-10% HNO3 for 30 min (remember to flush with H2O before and after HNO3, and also to remove column).
-------------------------------------------------------------------------------------------------------
By Anonymous on Friday, June 18, 2004 - 07:18 pm:
Quaternary amines like betaine and carnitine can be run on a silica column with aqueous mobile phases.
Dear HPLC professionals:
I have had a problem with bad tailing for a while with our system, and I couldn't solve it.
Here is the description of the system: It consists of a UV detector, an autosampler, 2 pumps, one controller, a PC, and everything is controlled by the software titled "Class VP."
I have tried to reduce injection volume; replace all new fittings, and filter; replace some shorter tubing; flush with IPA; replace column (new vendors) and other tricks, but all did not work.
I have checked the specifications of the system, and they are:
Sampling loop: 50 uL.
Dead volume (ave from my experiments): 90 uL - this is high
Detector cell volume: 8 uL.
Cell path length: 10 mm.
Outlet tubing is 0.13 x 600 (OD is 1/16 in.)
Any suggestion is welcome. I am anxious to try other approaches, to move forward.
Thank you.
-------------------------------------------------------------------------------------------------------
By R C on Wednesday, June 2, 2004 - 06:43 am:
You've done lots of things so far but we need more specific information to help you. What is the sample -- if proprietary, what is the general class of compound? What is the column -- manufacturer, model, dimensions and it's chemistry? What is your mobile phase -- and how is it prepared?
Basically, we'll use this information to see how sound your method is. To give the worst case scenario, if you're analyzing a compound with multiple basic groups, on an old fashioned ODS column, using poor buffers -- it will take a lot of work to get results acceptable to you again.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Wednesday, June 2, 2004 - 07:20 am:
What do you call bad tailing?
Your column to the detector connection is 600mm long??
-------------------------------------------------------------------------------------------------------
By Henrik Vogelius on Wednesday, June 2, 2004 - 08:15 am:
Is the problem just arrived or has it alwayes been there, is it the first time your application run on this system.
Try to change the gradient a bit and also wash with HNO3 for 30 min, (remember to flush with H2O before and after HNO3, and also to remove column)
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Wednesday, June 2, 2004 - 01:05 pm:
What type of tubing are you using? Are you cutting the tubing yourself? If so, what method are you using? Often tailing can be created by tubing which is not cut squarely. If you are using stainless steel tubing then you probably are already using precut tubing and this shouldn't be an issue. But if you are using PEEK tubing, one has to be mindful of the cutting method in order to assure properly cut tubing is obtained. Upchurch makes a tubing cutter that is used as a rotary cutter which will generally make good-quality cuts. To cut tubing using this tubing cutter, one you should rotate the tubing cutter around the tube 3-4 times and then snap the tubing at the point where the tubing was scored by the tubing cutter. Generally, it's not a bad idea to inspect the end of the tubing using an inspection microscope to make sure there weren't any significant defects introduced into the tube ends, as even with this method there can sometimes be defects.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Wednesday, June 2, 2004 - 04:04 pm:
How do you know that it is not the column? What column are you using? Also, you can check the width of the peak by injecting a sample without the column. With a extra-column dead volume of 90 microL, I expect the peak width (at 4.4% of the height) to be also around 90 microL, maybe just a bit better. If it is much worse, see how the injecotr is plumbed (I assume that you are not injecting the full 50 microL of the injection loop???).
-------------------------------------------------------------------------------------------------------
By Anonymous on Friday, June 4, 2004 - 06:01 am:
How about getting some of the standard that the column manufacturer uses to test their column (ususally a test sheet accompanies a column) and trying to repeat their run... If that works, then it is most likely a method development issue.. if not then perhaps theres a problem with your column or system.
-------------------------------------------------------------------------------------------------------
By DR on Friday, June 4, 2004 - 06:46 am:
More details are needed, as others have stated. 90µL is not bad for dwell volume, especially in a Shimadzu. Modular ones typically run ~2mL for modular high pressure binary gradient systems.
-------------------------------------------------------------------------------------------------------
By AlfredL on Wednesday, June 9, 2004 - 09:21 am:
Dear all professionals:
Thank you for all your responses. I will try to go back in time, and provide some more info.
The name of the compound is Oxymetazoline HCl (it has a monograph in USP 27). The mobile phase is MeOH:H2O:NaAc:HAc = 40:46:10:4.
We usually analyze it using a Supelco SCX column (4.6 x 250mm) on an older HPLC system (using a separate detector, and integrator). We also have some tailing problems on this one, especially when the column is almost dead. The sampling loop of this old system is 200uL.
We would like to transfer the method to the newer Shimadzu, which is controlled by a PC. But we observe problems.
Supelco SCX column (4.6 x 250) gave peak tailing. T ~ 2.4-3.
Phenomenex SCX (4.6 x 250) gave high back pressure.
Waters Spherisorb SCX (4.6 x 150) gave good peak shape at the beginning, but later on peak height decreased, and bad tailing (T > 3).
Since I would like to validate the method with a SCX column, I have tried several things, but none is satisfactory. (This is not a comprehensive list).
1.Reduce sample injection: I reduced from 20 to 10, to 3-4 uL (on the old system, I would inject 20 uL as the SOP).
2.Flush column and check for aging: The Waters SCX column was new; nonetheless, I have tried to regenerate using several solvents (as recommended by mfg), but the peak height was only improved for a few runs right after the regeneration.
3.Check fittings and connections: I replaced all the old PEEK fittings. Also, replaced the filter. No change.
4.Check contamination: I flushed the system with IPA, and then switched back to mp. No improvement.
5.Check for dead volume: I ran the Shimadzu system without column; water as mp at low flow rate (0.2 ml/min), and inject Acetone to measured the dead volume. Ave. was 90 uL. The peak had bad tailing also (T =2.6-3.4).
6.Check for secondary interactions. I believed that there could be some absorption on to the column due to uncapped Si, as well as some affinity due to the metal ions (present in the silica support).
I had to search several vendors to find low-cost SCX column with gel support (instead of Si base). I found two: Hamilton and Higgins.
a.The Hamilton column (3.2 x 150 ?) gave a chromatogram with no clear baseline, so I assumed that it is because of the pore size (100A). [A similar situation could be found on pp. 240 in the HPLC book by J. Swadesh, ISBN 0849300037]. Also, this column is made to analyze metal ions chiefly.
b.The Higgins column (3.2 x 20) needs a holder, but it reduced run time significantly. Peak tailing is still 2.1-2.2 > 2.0.
I then believed that the SCX columns (especially the Higgins) might not be the problem. I had to look into the system.
7.Check for plumbing: I looked into the manual, and found that the mfg installed tubing with ID 0.13mm (very long outlet tubing indeed). Visual inspection told me that it could be 0.3 or 0.5 mm ID (but the Errata of the manual stated that it is 0.3 mm ID). I decided to order precut tubings with ID 0.127 mm to replace current outlet tube, and more if I could (from inj. Point to 6-position valve).
8. One last thing I should have mentioned is that I have run a C8 column, with the premixed test solution as sample. The chromatogram was not similar to the mfg’s, and I observed peak tailing as well.
Since I am waiting for the arrival of the tubes, I cannot predict if it will solve our problem. But I will post the outcome.
Thanks for reading this long list.
Alfred.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Wednesday, June 9, 2004 - 04:55 pm:
What was the peak width of your acetone injections without column (measured at 4.4% of the peak height)?
-------------------------------------------------------------------------------------------------------
By DR on Thursday, June 10, 2004 - 01:02 pm:
I've used SCX columns in the past - Mine always had a nasty habit of voiding at the inlet. This quickly leads to profound tailing, splitting etc. I suggest that you get a cartridge guard column and a new analytical column. Inspect the guard column regularly (or keep close tabs on its back pressure w/o the analytical column in place) to make sure you have some spare packing between the pumps and the analytical column.
good luck,
DR
-------------------------------------------------------------------------------------------------------
By AlfredL on Thursday, June 10, 2004 - 03:56 pm:
Peak width of the Acetone peak at 4.4% peak height is 0.2428 min. (ave.), or equiv. to 48.6 uL.
-------------------------------------------------------------------------------------------------------
By AlfredL on Saturday, June 12, 2004 - 08:46 am:
I have replaced the outlet tubing (it has smaller ID and is shorter), but the tailing is not improved. Checking the system w/o column by injecting Acetone (mp = water), the tailing is still observed (but less than before).
What should I try next?
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Saturday, June 12, 2004 - 03:50 pm:
Based on all of your comments above, I think most likely the tailing you are observing is due to secondary interactions between your analyte and contaminants on your stationary phase. Generally, the effect of "plumbing" problems should be most pronounced for early using peaks. Even if you are observing an asymmetry value of 2.6-3.4 without a column, an analyte with a decent k' should exhibit significantly lower asymmetry values whereas you are sometimes observing comparable asymmetry even when the column is present. Since you report that under some conditions (at least initially) you observe respectable asymmetry, this would seem to suggest that the problem with tailing is related either to your injection solvent, your column or some sort of contaminant which strongly interacts with your analyte. Of these choices, I can't rule out the possibility that your problem is sample solvent related since you haven't revealed the specifics of your sample solvent. But the fact that you have seen the problem on different columns to varying extents suggests at least part of the problem is column related. The fact that the Waters column gave good peak shape it at least initially also supports that view. If you want to unambiguously sort out whether the problem is sample related, "plumbing related" or due to a column contamination problem. Try the following:
1. Make up your sample with as little solvent as possible and minimize the ionic strength of your sample.
2. Observe the effect of increasing retention on analyte peak symmetry by progressively reducing the ionic strength of your mobile phase.
If changing your sample solvent has a major impact on peak tailing than the problem is probably related to your sample solvent. Ideally, in liquid chromatography the sample should be dissolved in an "eluent" with lower elution strength than your mobile phase. Under these conditions, analyte focusing will occur during the loading step, minimizing peak distortion caused by "plumbing problems" in your injection system.
If changing the sample solvent has no effect, then varying the retention of your analyte well I you to ascertain as to whether or not the problem is caused by extra column affects (in this case the asymmetry will improve as the retention time increases) or caused by stationary phase effects (in this case the asymmetry will either stay the same or get worse as retention increases). If you observe the asymmetry increasing as the retention time increases, this would seem to narrow the problem down to either your stationary phase itself or some kind of stationary phase contamination. One possible contamination of a SCX column which might cause these effects for your analyte is iron contamination. This might explain why you observed acceptable symmetry initially on the Waters column but later saw degradation of peak shape. To test the metal contamination hypothesis, I would recommend treating the Waters column with 0.1 molar oxalate (pH 2.5) overnight. If you see improve peak shape after this treatment you can be fairly sure the problem is metal contamination related. In this case you might want to install a chelating trap column ahead of the injection valve to eliminate metals arising from system contamination.
-------------------------------------------------------------------------------------------------------
By AllsepTech on Sunday, June 13, 2004 - 07:11 am:
You might also consider our approach to tailing problems. We have recently developed a method for oxymetazoline with symmetry very close to 1.0. There is also a comparison with two leading brands of columns. Check this link, I think this will help:
http://allsep.com/makeChr.php?chr=Chr_065
-------------------------------------------------------------------------------------------------------
By AllsepTech on Sunday, June 13, 2004 - 08:16 am:
Oxymetazoline is a very strong base and tailing is probably caused by secondary interactions with residual silanol groups. If you are using end-capped column this tailing is less pronounced at the beginning, but during the life time of the column the endcapping is slowly coming off and you will observe more tailing. Non-endcapped columns will give you bad shape even at the beginning. We are using a different approach, our stationary phases do not have endcapping instead the silanol interactions are completely masked by our stationary phase. Even with strong bases like oxymetazoline and quaternary amines you will observe a symmetrical peak:
http://allsep.com/makeCmp.php?cmp=Cmp_114
http://allsep.com/makeCmp.php?cmp=Cmp_110
-------------------------------------------------------------------------------------------------------
By Anonymous on Monday, June 14, 2004 - 12:48 pm:
I am trying to analize betaine and carnitine but have a very bad peak shape using XTeraa column (it slghtly better on Polaris and Zorbax colomns, but still not acceptable by my internal requirements). Both compounds are internal quaternary amines. I tried different ration AC-Water and MeOH-water with buffers, but still have a triangle peak. How can I improve peak shape for this compounds?
-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, June 17, 2004 - 07:59 am:
You still got tailing w/o column, and therefore I still think you should clean the system with 5-10% HNO3 for 30 min (remember to flush with H2O before and after HNO3, and also to remove column).
-------------------------------------------------------------------------------------------------------
By Anonymous on Friday, June 18, 2004 - 07:18 pm:
Quaternary amines like betaine and carnitine can be run on a silica column with aqueous mobile phases.
