Interfering peaks

[sstarkey]

Hi, I have been running a Benzo screen on my LC/MS and out of the blue have seen the exact same contamination peaks. I have by passed the degasser, cleaned everything, tried new LC/MS grade water and ACN, by passed the injector, column switching valve, column heaters, everything, and I still have these contaminates. I also have the problem of the peaks co-eluting with my Benzo compounds which ruins the method. I have tried different Formic acid which reduced the contaminate but it is still present.
I have tried the gradient with excess equib. Time and the peaks do increase in size as the time increases. I have switched water sources, pump lines, degassers, and switched the solvents from channel A and B. The contaminates are still there.

The only thing that alters the contaminates is when I remove the guard column the "peaks" disappear. When I install the guard whether it is brand new or used, the peaks return. I am desperate for a solution because without this method we are missing some Benzos and having to spend money sending out the samples to an external Lab.

Has anyone been able to locate the source of the contaminate? I have spent months trying to fix the problem and can't figure it out. Any help/guidance/solutions would be much appreciated.


My LC system is an Agilent 1100 with degasser, MS is an Agilent MSD. The mobile phase is 0.1% Formic acid in H20: 0.1% Formic acid in ACN. The gradient is 10 - 40% B in 9 min.

[tom jupille]

when I remove the guard column the "peaks" disappear

How often do you have to change the guard column in routine use? Could you simply leave the guard column off?

[sstarkey]

The guard column is usually changed when the pressure increases about 50 bar higher then with a new guard column. I work in a forensic toxicology lab so we want to validate our method with a guard column to try and extend the life of our columns. We use a liquid-liquid extraction but sometimes there are some really bad de-comp samples and we want protect the columns from the samples.

[tom jupille]

Columns are cheap; your time is valuable.

That said, the fact that the size of the artifacts increases with increasing equilibration time points to low-level contamination in the "A" mobile phase. Is the packing in the guard identical to that in the analytical column (if it's more retentive, for example, you might have garbage accumulate on it that might pass through your analytical column unretained)? If the guard *is* identical to the analytical column, then maybe there is some garbage already there. Perhaps a pre-wash with 100% organic before you install it will help?

[Jetjamnong]

sstarkey,
What about the glassware. Do you clean the glassware properly? e.g. all test tubes for LLE, LC vials, micro-volume insertes! or even the solvent you use for reconstitute.
In my experience about contaminate, some time from the following sources.
- Carry over from high concentration sample (Wash the injector needle internal and externally several time).
- Glassware contaminate (you shoud use all new glasswares).
- Solvent contaminate especially the reconstitute solution (Prepare new solvents).
- Residues of tablet, capsule, and /or any powders of evidence (Do not use the same lab area for prepare the sample from tablet/powder and biological fluids and separate them glasswares).

I have some question with my opinion promptly. What's type of your contaminate, baseline or peak(s) are the problem?
- if baseline is the main problem, maybe from mobile phase solvents, such as water, ACN, Buffer and some time it's from the ion source.
- if peak(s) present in solvent blank injection, maybe from previous sample(s), glasswares, solvents
- if peak(s) is occured, is it pass all criteria with RT, relative ion intensity at minimum 3 ions.

This is only my opinion and maybe it can help you.

Best Wishes
Jetjamnong

[sstarkey]

The guard columns are flushed with 100% MeOH before being flushed with the mobile phase. I agree that there is probably a contaminate in the mobile phase but I am not able to determine which component is the problem. Since many people suggest that it is always the water we purchased LC/MS grade water and tried that but the problem was still there. The only other ideas I had was that it was the formic acid but after reading the other users posts and he had the same type of contaminates but was using acetic acid I am doubtful that that caused the problem. I am still waiting to purchase a new bottle of ACN to see if maybe that is causing some type of contaminate.

The guard columns are the same type as the analytical column. It’s a Zorbax eclipse plus C18 2.1 x 50 mm, 3.5um column and a 2.1 x 12.5 mm, 5um guard column.

The solvent bottles get rotated every week and the Mobile phases get changed as well.

The solvent bottles are rinsed with hot water, IPA, and MeOH. The contaminate peaks appear when the gradient is run alone without an injection.

The main problem with these contaminate peaks is that the main ion, 287, interferes with the integration of the IS and therefore causes every cal and control to fail.

Our issue with “columns are cheap” is that we don’t really have a budget to stock pile columns. Our last method was validated on a column without a guard column and we were only averaging about 25-50 injections per a column and a column was over 700 dollars. With our new methods we were trying to save money and down time by using a guard column. So we have 2 frustrations the first being this unknown contaminate and the second being unable to use the method.

The contaminates are still present without the guard column the abundance is just too low to interfere with the method.



Is it possible the contaminate is coming from the ACN or the formic acid?

Is there any easier way to test for the contaminate source other then trial and error?


Any help would be appreciated!

[Jetjamnong]

Which Benzodiazepines given you with m/z 287?

[sstarkey]

Oxazepam and Diazepam

[Jetjamnong]

Yes, Oxazepam is m/z 287.0582 [M+H] but about Diazepam is m/z 285.0789 [M+H], could you tell me about RT of those compouds.
Oxazepam, RT = ?
Diazepam, RT = ?
Interference RT = ?

and some information about the LC/MS type, Single Quad or TOF? maybe it's not contaminates but done wrong in some points of operation.

[Peter Skelton]

The guard columns are flushed with 100% MeOH before being flushed with the mobile phase. I agree that there is probably a contaminate in the mobile phase but I am not able to determine which component is the problem. Since many people suggest that it is always the water we purchased LC/MS grade water and tried that but the problem was still there. The only other ideas I had was that it was the formic acid but after reading the other users posts and he had the same type of contaminates but was using acetic acid I am doubtful that that caused the problem.

Well try changing it anyway. It definitely seems that you have a mobile phase contamination issue and as you've already changed the water then the acid is the other prime suspect.

[sstarkey]

We use the M +3 ion as a qual ion, so for Diazepam its 287. The retention times shift with the gradient but traditionally Oxazepam is ~7min, Diazepam is ~8min. The contaminate is a double peak that comes out between the 7 and 8 min. The contaminate fragments to 269 and 228.

[Jetjamnong]

Do not use isotope ion or adduct ion in forensic work (SOFT 2006), thus M+3 (Cl isotope) not good for both quant and qual ion. Try to look at m/z 154, 193, 222 or 257 for Diazepam and 104, 241 and 269 for Oxazepam.