I'm working on a method development for several compounds in plasma, by LC/MS/MS. These are MTBE extracts, blown down and reconstituted in an LC friendly solvent. LC is gradient elution on 2.1 x 50 mm C18, aqueous / methanol with 0.1% formic acid in each. Right now I'm running check standards, which are merely solvent standards, interspersed with the samples. The problem is, after a while, the check standards will fail on the low side for the later eluting analytes only. The earlier eluting analytes seem to be stable.
My thinking is that there are strongly retained components building up, then eluting in subsequent runs, suppressing the ionization of the later eluting compounds. Any other suggestions as to what is happening will be welcomed. Since the earlier eluting compounds, including internal standard, don't seem to be affected, I don't think the problem is simple fouling of the ion source.
Assuming that the problem is with strongly retained components, I'd like to do a wash, and worry that if these are lipids or something, methanol might not be sufficient to clean the column. I am already doing a 4 min. wash of 97% MeOH at the end, and at my flow rate of 0.35 mL/min, this should be 14 column volumes. So I am thinking a stronger solvent is in order, and am looking for suggestions. I'd like to keep this a fast method. Would pure ACN be much better for removing lipids or strongly retained proteins? What about something like acetone? I've seen that one used for chromatographing lipids before. UV background is not an issue here. Whatever I end up using, I'll be diverting the column wash to waste rather than blowing it into the ion source. Thanks for any help!
By Uwe Neue on Friday, April 30, 2004 - 03:23 pm:
Acetonitrile is a stronger eluent than methanol, and acetone is stronger than acetonitrile. If the problem is indeed what you suspect, an acetone wash might help.
How are you doing the quantitation, peak area or peak height?
By MG on Friday, April 30, 2004 - 03:26 pm:
I'm quantitating by area. The peak shape and retention times do not appear to change with the later check standards, but the areas and heights are diminished.
By HW Mueller on Monday, May 3, 2004 - 12:13 am:
As mentioned earlier, if proteins are a problem, you could be literally cementing them to the column with pure MeOH or ACN. To get rid of such proteins you need chaotropic salt solutions, urea or detergents, maybe even a reducing agent like dithiothreitol.
Maybe you can keep the proteins in solution by some choice of volatile chaotropic substances in your mobile phase? I thought that the protein MS people had solved these problems.
By Anonymous on Monday, May 3, 2004 - 08:30 pm:
My experience is HPLC using UV and fluorescence detector only, not MS. But I use C18 column and my samples are plasma or serum like you.
In my experience, after I finished my work, I have to wash column with 5%acetronitrile : 95%water (or 100% water can be used) first, then wash with 95% acetronitrile, many dirty compounds came out. Washing C18 with water first is likely to "trigger" something in column. If I wash with 95% acetronitrile suddenly the dirty compounds are not likely to be eluted from column.
I'm just the beginner in HPLC, this is only my experience and what I think. Sorry if you already knew or it's not involved with LC/MS/MS.
By HW Mueller on Tuesday, May 4, 2004 - 12:29 am:
It might be in order to give a brusque summary of what worked for us:
Precipitate out some of the plasma/serum proteins with Na2SO4 (even when restricted access columns were used, as some samples were from catabolic patients with strange proteins). With mobile phases between ~30 to 60% organic (especially alcohols) most proteins appear to be washed out of the column at ~tm. When this is the case one can go to 100% organic immediatly after the last run (should be done at every working days end). After the lipids and other hydrophobic substances are washed out by the 100% org. the column is put back on mobile phase (H2O+org.) over night, to prevent a jump of backpressure the next morning (this backpressure increase with 100% org. was taken as evidence that some proteins were left on columns). In time (weeks) there may be an increase of pressure anyway, which can usually be corrected with Li-dodecylsulfate/dithiothreitol (some details in J Chrom B, 678, 137 (1996)). Later it appeared that the detergent/reductant should be used only as a last resort. Chaotropic salts or urea seem to be a gentler method to remove obstinate proteins.
By MG on Tuesday, May 4, 2004 - 07:25 am:
Thanks everyone for the answers. I had one more injection available from these extracts, before more would have to be prepared. I wanted something that could be done quickly and as part of the online LC/MS run, so I decided to try the acetone wash (gradient to 100% acetone after compounds of interest had eluted). Let me tell you, it greatly improved the stability of the method. The problem appears to be solved.
HW Mueller, hopefully I am washing off the proteins then in my initial conditions of ~ 50% methanol. I don't see any backpressure increases at this time, so hopefully I am not cementing any proteins to the column.
By HW Mueller on Wednesday, May 5, 2004 - 01:31 am:
Unless you have some of these "funny" proteins it will take quite a while to give problems, since most proteins will be washed out with your mobile phase.
I forgot to mention that I also add a guard column or at least a pre-frit when proteins are involved. The first step after trouble is really to reverse flow of the guard or frit (alone). That usually largely restores flow characteristics. When the column is involved it is usually also subjected to reverse flow before the cleaning, mentioned above, is done (the evening methanol wash is always performed).
By MG on Wednesday, May 5, 2004 - 08:12 am:
Yes, I *always* use a guard column. And I've been known to backflush guard columns to get a little more use out of them. Sometimes it works and sometimes it doesn't. :-)
Thanks Uwe, HW, and everyone else for your help with this problem!
By Anonymous on Thursday, May 20, 2004 - 07:24 am:
I have a similar problem with proteins - I think - of a tissue culture broth/extract. I am analyzing for low levels of sugars and get many small peaks that overlap with the expected sugar(s) peak(s). Waters suggested a developmental kit "C" that has two SPE cartridge media (anion and cation) that will tie up the proteins and let the sigars through. I am awaiting the kit. You might want to contact them to see if they can help you out. "..cementing proteins to a c18 column with acetonitrile..." I am not using a c18 column, but a supelcogel h column since I am also analyzing for lactic acid and a few other organic acids. What will proteins do to a supelcogel h column?
By srinujayanthi on Wednesday, June 9, 2004 - 07:01 am:
i am having problem in washing the c18 shimadzu column
By Anirban Roy Chowdhury on Thursday, June 10, 2004 - 08:15 am:
I want to know why you need to wash Reverse phase columms with organic phase after washing with water.Reverse phase columms are non polar right so water that may be left over should not cause any interaction.
By Chris Pohl on Saturday, June 12, 2004 - 02:18 pm:
The answer to your question depends somewhat upon the specifics of what preceded the above mentioned wash steps. Frequently, when switching from a buffered mobile phase to a high solvent content mobile phase, it is preferred to rinse the column with water in order to remove buffer components. Otherwise, some buffer components may precipitate when brought in contact with a high solvent content mobile phase.
On the other hand, if you've been using a 100% aqueous eluent, rinsing with a high solvent content mobile phase will re-wet the stationary phase in case dewetting has occurred. If you're interested in learning more about the specifics of how to operate without having to worry about dewetting, you will find numerous threads on this topic in the archives. In general, the best solution is to use stationary phases resistant to this problem. One class of reversed phase materials generally resistant to dewetting is polar embedded phases.