increased back pressure

[harringg]

Quick background:

I'm running a 2.1x100mm C18 column, 2.1x10mm guard column, flow 0.2 ml/min, 60-40 A:B to 0:100 A:B in ten minutes, flush 5 minutes at 100 B, back to starting conditions long enough to equilibrate, A:HPLC grade (bottled) water, B:ACN. Currently have about 700 injections on column, detector is Qtof, ESI-, MS-MS w/fragmentation to quantify. This is the setup we have been running for the last month, samples are groundwater cleaned up via SPE and in 50:50 H2O:ACN. Guard colum was changed about 300 injections ago.

Last week (and for several weeks prior) back pressure has been ~1200 PSI, areas of standards and IS have been fairly consistent (~20 for IS). The last few days, the back pressure has climbed to as high as 1500 and yesterday it was only 1400. Delta PSI is about 15. The areas of our standards and IS have increased 3x in area from yesterday to today. My question is: as a result of "stuff" getting on the column, is the compound binding better, and more is getting held and then released? It seems so drastic for it increase 3x better sensitivity (IS = 60 area units). Retention time is the same for both days and hasn't changed when PSI was 1200 vs the now 1500. I'll cross post in Hyphenated techniques, but it seems like an LC issue more than MS.

Thanks,
Grant H.

[Mark Tracy]

Your hypotheses is inventive, but likely wrong. I don't think that your column has been eating 2/3 of the sample for the last 700 injections, and now is suddenly overflowing. If sample adsorption were the problem, you would have suffered other problems long before this (bad calibrations, carryover-like effects, bad peak shapes). "Deactivating" the column would also produce changes to the retention time and/or peak shape. But this is all easy to test for: install your spare column.

There are really only three possibilities for the increased signal: dilution error, autosampler malfunction, or better ionization efficiency. The first two possibilities are pretty easy to eliminate. One guess is that your nebulizer is partially blocked in a way that raises the backpressure and alters the spray pattern. Do you know how much backpressure your nebulizer makes when it is operating normally? Are you seeing similar increases in non-target ions like solvent clusters?

[harringg]

>There are really only three possibilities for the increased signal: dilution error,

Nope. Sample came from same vial today as yesterday (stored at 4C overnight).

>autosampler malfunction,

And Nope. 10 ul injection in 25 ul syringe, at most the error would be 2.5x

>or better ionization efficiency.

Could be.

>The first two possibilities are pretty easy to eliminate. One guess is that >your nebulizer is partially blocked in a way that raises the backpressure >and alters the spray pattern. Do you know how much backpressure your >nebulizer makes when it is operating normally? Are you seeing similar >increases in non-target ions like solvent clusters?

If by nebulizer, you are refering to the capillary tubing of the ESI source, the spray pattern "looks" normal. I'm running MS-MS and have the Q1 set to only allow M and M+1 isotope through and then fragmenting, so there aren't any to check. Thank you for the response, maybe a better question would be (since I'm still fairly new to LC-MSMS) using only ACN:H2O as mobile phase and "clean" samples, what would cause the LC backpressure to increase 300 PSI in a matter of a week or less, when it's been running constant for several months? Also, we keep the LC constantly flowing with solvent, so it's not shut off/turned on to allow for phase collapse, and when the column has been stored, it's been in 80:20 ACN:H2O and capped.

Thanks again,
Grant H.

[HW Mueller]

Do you know that your flow rate stayed constant?

[harringg]

>Do you know that your flow rate stayed constant?

Yes, because I've been watching it so closely. I'm running the samples from a sample list where the LC parameters are downloaded to the LC. The flow is set to 0.2 ml/min and the LC (2695 Alliance) system is set to a 25 ul stroke length, which will only allow at max ~0.5 ml/min. Just trying to figure out what would cause backpressure to rise (with everything else being constant: tubing, tube length, flow, mobile phase [aqueous replaced every two days], etc...) yet the RT stays the same, 1200 psi:5.26, 1400 psi: 5.31, 1500 psi:5.28, but overall area inceases. It could an ioniztion issue which could be related to the reason why the column backpressure is increasing. I guess I'm new enough to not know when a column needs replacing. As long as all is fine in terms of linear quanitification, RT are constant day-to-day, should I just ignore the increased backpressure until it gets so high, and the system trips? Then replace it? I know I could backflush it, which I may do too. Sorry for the detailed descriptions, just trying get a better idea of LC column use from those who are far more knowledgable than me.

Thanks,
Grant H.

[Mark Tracy]

Grant,
I agree that your pump is OK. You have eliminated two of the possibilities related to the sample.

Columns do wear out with time, and high pressure is one common symptom. One cause is colloidal particles from your samples; 0.45 micron filtration may not be enough. (Story time: I once saw a high-pressure column that had been used to test water from the Red River. The column turned red from the clay particles.)

"Bed collapse" is a mechanical degradation of the silica support that may cause high pressure. Usually it is accompanied by other severe symptoms.

So, as long as this column passes all its system suitability tests, keep using it. Just be prepared to replace it.

I hope you don't mind a bit of unsolicited advise. An unbuffered mobile phase has no control over pH and ionic strength, and is at the mercy of trace contaminants. Both factors do influence the ionization process in the ESI source. For your next application, design the mobile phase to put you in control of those factors.

[harringg]

Thanks for the advice. I'm looking at data from last night, and even though the column is still giving high backpressure, the data looks solid, so I'll carry forward. Thanks for all the usefull insight.

Thanks,
Grant

[HW Mueller]

The pump setting doesn´t say much about flow, liquids are compressible (to be honest, I have a bit of a conceptual problem here), also there could be an undetected leak, flow should be measured. But, since in your case the retention times stayed constant your flow must have stayed constant. Also, a 3x change in flow, which could account for your area increase, should be hard to get. Your problem lies elsewhere. Also, I have never seen an increase in peak area due to an increase in backpressure alone (don´t see a theoretical reason either). I just wonder, did your peak hight increase 3x as well?