Column Contamination

[AMacD]

Hello,

I am developing a drug confirmation method (in urine) on an Agilent 6400 LC/MS/MS. The method is for benzodiazepines and opioids and there are hydrolysis, extraction, and filtration procedures. I have noticed recently a hump (~1000 counts) near the end of the run (9 min), which I think is a result of a few benzos sticking to the column (oxazepam, temazepam, lorazepam, as I can tell from the MRM transitions). It is interfering with my recovery study (the recovery for my six replicates gradually increases for the compounds mentioned above). I have increased the time at the end of the run for which I hold at 100% acetonitrile. I tried rinsing with 100% organic three days ago, but this didn't help. Over the weekend I tried washing with 20 cvs (column volumes) ACN, 5 cvs IPA, 20 CVs hexane, 5 cvs IPA, and back to 20 cvs ACN, but to no avail. I backflushed the column (Zorbax C-18 2.1 x 100mm 1.8 um) this morning at 0.1 mL/min with ACN. When this didn't work I backflushed with IPA, and this still hasn't gotten rid of the hump. I run at 60 degrees C. Does anyone know of a wash program that might remove these benzodiazepines from this column? I will probably have to start using a new column, but I thought I would check to see if anyone had any suggestions first. Any help you can provide would be greatly appreciated.

[Uwe Neue]

1. What are your mobile phase conditions?
2. Why do you think that the diazepams are sticking to your column if you get peaks from them that elute somewhere else? Have you considered that they may stick to the injector?

[AMacD]

I am using 0.1% acetic acid as my aqueous phase and acetonitrile as my organic. The method starts with 100% acetic acid and has a gradient over 8 min, going up to 100% organic.

There is no contamination peak present when the column is not in the flow path.

[AMacD]

In response to your second question, the benzos elute at the same retention time as this hump. I think they have been building up over time with use of the column. It may be a result of not holding at 100% organic at the end of my run for a sufficient time, which I have now changed in my method.

[Uwe Neue]

I have nearly never seen contamination peaks from small molecules that come from the column. This, together with your failure to fix the problem, is why I am skeptical that the problem is with the column. Any tee in your flow path?

[AMacD]

No, I am not using a tee. I think you are right with it not coming from the column. I tried a new column this morning and injected a blank and am still seeing the peak. I cleaned the injection valve and rotor seal, made fresh mobile phase, but the peak was still there. It is definitely worse after an injection of a high concentration sample (500 ng/mL), but does not decrease with subsequent blank injections. I am not sure where else the benzos could be sticking. If you have any other suggestions, they would be greatly appreciated.

[tom jupille]

Run the "three blank gradients test" as described on our web site:
http://www.lcresources.com/resources/TSWiz/hs400.htm

That will tell you whether the problem is coming from your "A" mobile phase reservoir or from elsewhere in the system.

[AMacD]

Thank you, Tom. I ran the three blank gradient test this morning and the area of the peak in the third run is about 2.5 times greater than the area in the second. I just re-prepared my aqueous mobile phase using a different bottle of glacial acetic acid (since I tried preparing new mobile phase yesterday in a fresh bottle and the peak was still there) and am flushing the system now. I really appreciate your help.

[tom jupille]

You're welcome!

But now comes the hard part: tracking down the source of the contamination.

[AMacD]

Yes, tracking down this contamination is becoming quite a headache. My aqueous mobile phase is 0.1% acetic acid and I have tried preparing it from two different bottles of glacial acetic acid, flushing the line with IPA, trying formic acid, using the second line we have for our aqueous solvents, and preparing the mobile phase from two different sources of water and the peak is still there in my no injection runs. If you have any other suggestions on something else I could try, they would be much appreciated. Thank you.

[danko]

Hi AMacD,

Some people filter their mobile phase. Do you? If yes, skip the filtration – just to examine the possibility of contamination from the filters or the filtration equipment.

Another source could be the reservoirs. I’ve seen it several times – typically some detergent residues from the washing procedure. To examine; rinse the reservoirs several times with small portions of mobile phase before transferring the bulk of it in the reservoirs

Best Regards

[tom jupille]

If you have been working with your compounds for a while, it is possible that your lab has become contaminated and they are getting into *everything*.

"Plan B" is to remove the junk from the A solvent before it gets to the column.

If you have a "high pressure mixing" (aka "two pump") system, then you can put a C18 guard cartridge in the line between the "A" pump and the mixer. Since this never sees the strong solvent, any junk in the weak solvent will get trapped and retained.

If you have a "low pressure mixing" (aka "one pump") system, then the guard cartridge trick won't work, but people have reported good results (sometimes) by filtering the "A" solvent through an SPE (solid phase extraction) filter disk. Do a search of the Forum for "Empore" and you'll get a feel for the details.

[AMacD]

Danko: No, we have a water filtration system and we buy the HPLC grade solvents. I do have inlet filters in my solvent bottles, which I sonicated, but still see the peak. Yes, I usually make a habit of rinsing my bottles with whatever I am putting in.

Tom: I think I will have to try the guard cartridge. I performed the time consuming nitric acid wash yesterday (didn't work) and cleaned the MS capillary today in case there was something in there, but the peak is still there. Hopefully the guard cartridge will work.

Thank you both for your help. It is much appreciated.