I am using a ThermoSep 3200 pump with a gradient system as follows: A: 20/80 50 mM KH2PO4, pH 2.51/AcN, B: 90% AcN. Everytime I try to run this gradient (0 -20 in 20 minutes, then to 100% B in 5 minutes), the pump seems to clog up and I have to purge the system. Is there prepcipitation happening? This pump has traditionally run only TFA gradients in the past, and introduction of salt buffers is "new" to it. When I reintroduce a TFA solution, the pump is fine. Any help or suggestions? I am trying to run a barbituate assay and time is getting short to complete the project.

Your problem is caused by precipitation of the buffer. Phosphate salts are not all that soluble in polar organics like ACN, MeOH, etc.

I would suggest you try one or both of the following :

1) Reduce the proportion of ACN in solvent to about 60% or thereabouts.
2) Use a gentler gradient. I think 20 to 100% B in five minutes is a bit too steep for your system.

Also, a pH of 2.5 is too low for most HPLC systems, in particular the pump seals and the column. I assume you are using a C18 column.

Nowadays you get ultra high mol.wt pump seals (UHMW) for such applications. These are generally more resistant to extreme pH.

Further, your c18 column can't take anything outside a range of pH 3-11. In case you have no choice but to use only pH2.5, then you need to use a saturator column before your main column.

This will prevent your main column from dissolving at that pH.

Of course, all these changes will have an effect on retention times and/or resolution.

In which case, you can consider using an ion-pair technique to separate barbiturates. If you need any info. on this, do let me know, perhaps I can help.

regards,

s.k. srinivas
stimulus/bangalore/india

Some comments on pH and column stability.

"Further, your c18 column can't take anything outside a range of pH 3-11. In case you have no
choice but to use only pH2.5, then you need to use a saturator column before your main
column. This will prevent your main column from dissolving at that pH."

Well, I have to take issue with what s.k. srinivas stated above. I think you recieved some incorrect information. If a C18 (or other phase) column is SILICA-BASED, it should not be operated outside the range of 2-7.5. Although there are some newer brand columns on the market that clain to be stable at high pH. At low pH's the bonded phase will be cleaved (not dissolved) and most often, RT's will decrease. Although, RT may increase for polars due to interaction with silanols. Columns don't last as long at low pH (C18's typically last longer than C4's), but we use them all the time at low pH's when we run protein/peptide assays using the common MP of ACN:TFA. TFA has a very low pKa. At high pH's (above 7.5) the silica will begin to dissolve. Therefore, you don't need to use a saturator column at low pH!

With respect to your main problem of precipitation (pptn) of phosphate buffer. You can reduce the concentration of the buffer (50mM to 25mM) and pptn will be less likely as you increase your ACN. But even at 25mM, you won't be able to go to 90 or 100% ACN. If in doubt, hand mix buffer and organic to see if pptn will occur before having your system do the mixing. Phosphate buffer is more soluble in MEOH, therefore, you can go to higher conc of MEOH than ACN before pptn begins. Also, low pressure mixing systems can have problems with mixing phosphate buffers and ACN at amounts of about 55% of ACN or greater (due to demixing). Although, the problem will not be apparent at first. Good luck!

If I understand your MP correctly, you have a 20/80 buffer/ACN mixture in A, ACN in B, and your gradient is essentially 0-100% B. If this is correct, you have very little water in the system--only 20% at the most. This gradient is essentially 80-100% ACN. Is this what you intended?

To anonymous 9/10/99:
Yup, that's correct. It's a new application handed down from R&D, and I'm finding that it is EXTREMELY hard to work with. I haven't tried the run again on the same pump..I just finished fixing it from the last pptn problem, so don't want to go that route again. our applications dept really enjoys phosphate buffers and low pHs (not the first time I've had this problem!)

There are some advantages to low pH and phosphate, especially when dealing with basic samples. Phosphate is relatively transparent, so you can use detection to <200 nm, and low pH is now generally accepted to give less tailing of basic compounds. Acetonitrile as the organic solvent is also preferred, for reasons of low-UV detection and low pressure drop. Having said that, it is possible to live with these conditions and at the same time minimize problems with precipitation. Obvious steps are (a) minimize the buffer concentration and (b) keep the water content as high as possible. When working at pH 2-2.5, phosphate is an excellent buffer that does not require a high concentration; 10-25 mM should be maximum, and even lower is possible. In order to keep the water concentration high at every point in the system, try working with A- and B-solvents that are as close in %B as possible. Precipitation in the mixer can occur when 100% aqueous buffer mixes with 100% organic, even though the final concentration is low in organic.

Is there a good pH 7 buffer that is volatile for use in LC/MS? All the standard pH 7 buffers are involatile and do nasty things to LC/MS sources.