Hi everyone,

I have been running my LCMS system on a particular drug along with an internal standard for over two months and everything was fine until recently. My drug of interest suddenly began splitting into two peaks and it is reproducible. However, the internal standard remained as one peak. How can this be possible, for one compound to split into two peaks with the other compound remaining unaffected? I did not change anything in both my chromatographic and MS conditions. The splitting of peaks just came up suddenly and I just don't know how to go about solving this problem.

I would appreciate any help/advice that anyone can give me.

First thing, try a brand new column of same part number. Inject mobile phase only to see if baseline is flat, no extraneous peaks. Isolate the problem by changing only one thing at a time. Always go back to "new" to troubleshoot, if the change helps, you've got it, if it doesn't. switch back to original column, plumb in a manual injection valve temporarily instead of autosampler, etc. Good luck.

Hi P.

I hope you don't mind but I'm answering your e-mail here.

Sometimes in gradient methods if the injection gets split due to a partially blocked frit or an injector problem the first peak or early peaks can get split because they were eluting down the column at the begining of the gradient. Later peak(s) won't be split because they were not eluting down the column and the analyte was concentrated at the head of the column.

Is the split peak the first peak? Is the sample prep'd in mp? what is the mp? what kind of column? isocratic or gradient? class of drug? have you tried switching columns with the same sample/mp?

The more info you give us to work with the better your chances of getting a good idea.

Hi Tom,

I am using gradient method. The split peak is the second peak. I use liq-liq extraction for sample preparation. I redissolve the residue after concentration using a small amount of EtOAc (extracting solvent) and backextract it with the aqeous part of mobile phase which contains water, ammo.acetate, and formic acid. I then inject it into LCMS. The gradient composed of two diff solvents, one containing the NH4Oac and formic acid in acetonitrile and the other is in water. I am at present using C18 column (2.1mmx50mm). The drug that I am testing is basic. I am now trying to switch column to see if there¡¦s any difference. Will inform you about results later.

Thanks

Hi Tom and Mr. (Ms?) anonymous,

I have tried switching columns but the split peak still appears. I hope you can give me more advice.

Thanks

Have you changed you supplier of amm.acetate or formic acid? What kind of water are you using (HPLC grade or ged. water)? Maybe there is something with you quality of your reagents??

Does this compound consist of a mixture of isomers? I've seen some compounds that consist of a mixture of diastereomers give one clean peak on a new C18 column. As the column ages and more silanol sites are exposed, the diastereomers start being resolved somewhat, giving a split peak.

all the reagents that I used were still from the same batch of stock. No change in supplier. I am using Waters' Milli-Q water. The compound that I am working on do not have isomer. I hope these info can help. Thanks for advice, hope there'll be more while I try working on it at the same time

How about a chemical reaction causing isomerization, ahead of the injection? Usually sudden changes like the one you describe can be due to subtle or not so subtle inadvertant changes. A similar problem is just dogging me (with ouabain, mentioned an aspect before).
Did you collect the peaks and reinject them? What happens if that is done?

You may have developed a void volume in your column--but you changed columns, to presumably a new one. Are you running a guard, or pre-column? If so, that is the most likely place for it to be. The column packing taking the first pressure 'hit' will crush, or move, or both, first. This problem has caused me much grief over the years.

The issue is a multipath problem. The early peak isn't split because it is flushed from the column prior to diffusion. The second peak is split because it has begun to migrate through the void volume while it is waiting to be eluted. If the pressure has gone up then your column is probably toast. Is the analyte basic enough to dissolve the silica once dissolved in the acid (solution pH)? If you run the assay everyday for two months then redevelopment is probably not indicated. If you have to keep running you will need to modify your gradient. You should look at the chromatography prior to the peak splitting. If the peak widths are the same for both peaks then your method was optimized. However, if the peak widths are wider for the later eluting peak you should consider increasing the slope of the gradient. This may allow you to continue your analysis for longer periods of time. Hope this helped.

Hi P.

Sorry I've been on vacation. Let us know if you got this figured out. I have not seen a void produce a split in a later peak but not in earlier peaks but I guess dan and Scott have. I am assuming your split peak is the same compound as detected by MS. Good luck.

Hi All,

I just want to inform you all that I finally found the culprit in my split peaks (it used to be just the second compound, then it got worse and both two compounds became split later). I used another guard column and analytical column but still the problem persisted. Regeneration of the column did not help either. Then today, I tried isolating the autosampler by using manual injection. The problem came from my autosampler after all. At last, I got my good old peaks back again!! Thank you Tom, Mr/Ms anonymous, HW Mueller, Dan and Scott Fredrickson for all your helpful advice. Welcome back Tom, I hope you had a lovely time during your vacation.

Best wishes to you all

Thanks for coming back and presenting the solution!

Hi HW,

I am grateful for people like you and the others for taking the time to share their knowledge about the technology and give advice to people needing solutions to their analysis problems. I hope that by presenting this small solution, it might also help anyone who might experience the same problem like mine (and hopefully needed exactly the same solution too!).

Best wishes

Hi All,

I'm back again because it seems my peaks are ok only at higher concentrations (with cps >2000). I don't have any problems with my internal standard (which comes out at around 5.8 mins) because it's count is >>>2000. However, with my compound of interest (which comes out at around 6.1 mins), I can still see split peaks coming out at 5.75 and 6.1 mins respectively if the cps <2000. I mentioned to you all in my second to the last message that I got my peaks back again after isolating the autosampler and used the manual injector instead. But my rejoice was shortlived because both peaks were only ok at higher cps, not at low cps like I mentioned above. My guess was that somehow, some dirt in part of the connections in the autosampler was the one that was causing the split peaks and the disconnection somehow removed the dirt a bit but not completely. I am thinking of replacing parts of the autosampler, hoping that things will improve. Please advice if you have some other ideas as to what's causing this split peak at low concentration/cps.

Thanks!

Are you absolutely sure that both peaks of your pairs are the SAME compound? Again, what does reinjection give?

Did you change the composition of your autosampler (AS) wash solution? If it was replaced and was too strong (i.e. high MeOH/ACN concentration) that could cause splitting.

Rule of thumb: prepare AS wash solution at about the same organic concentration as the MP. In other words, prepare the mobile phase without the salts. If all else fails, use 20% MeOH in water.

BASED UPON THE NUMBER OF COMMENTS THAT YOU HAVE RECIEVED WITHOUT A SOLUTION I THINK THAT YOU NEED TO BACK UP A LITTLE BIT. THE FACT THAT THIS IS A GRADIENT IS TROUBLING. EARLIER I ASKED IF BOTH PEAKS WERE SPLIT. ARE THE PEAKS AT 5.8 AND 6.1 MIN. OR ARE THESE THE SPLITS. LETS SIMPLIFY THE SITUATION. KEEP THE SYSTEM UNCHANGED AND INJECT A SOLUTION THAT YOU EXPECT TO SPLIT. IF THE PEAK(S) ARE STILL SPLITTING THAT IS GREAT. UPDATE THE GRADIENT TO RUN ISOCRATIC AT THE MOBILE PHASE COMPOSITION THAT IS THE SAME AS THAT CORRESPONDING TO 6.1 MINUTES. IF THE PEAK IS NO LONGER SPLIT THIS WOULD BE INDICIATIVE OF A MULTIPATH PROBLEM. THE CAUSE COULD BE EITHER THE COLUMN OR THE FITTINGS. THE CONCENTRATION DEPENDANCE WOULD MAKE ME THINK THAT THERE IS A VOID IN THE COUPLINGS. ALSO, THE FACT THAT THIS "JUST HAPPENED" MAKES THIS MOST LIKELY ALSO. CHECK ALL FITTINGS. BEFORE YOU CRACK INTO THE AUTOSAMPLER, MAKE SURE THAT THE LOW CPS SOLUTIONS DON'T SPLIT ON MANUAL INJECTION ALSO.