I am currently trying to develop a peptide seperation employing a water vs ACN gradient at 215nm and 0.05 AUFS using reverse phase chromatography. Mobile phase A consists of 0.1% TFA in DI water. Mobile phase B is a 0.08% TFA in 80% ACN. The gradient is developed over 40 minutes
from 0-100% B. During a blank gradient I see two sharp peaks corresponding to ~18 and 20% ACN. I use TFA in ampules and a high quality brand of ACN. I have tried new sources of both and still these peaks persist. They are sharp peaks and have the exact same retention times from run to run. Any ideas?

First, unless you have to, don't go to 100% aqueous -- the bonded phase tends to collapse. Start at 5%B if you can. In terms of extra peaks, I suspect these are coming from the water. Try to run the normal blank gradient then repeat with 5x the equilibration time (e.g. 25 min instead of 5). This should allow aqueous contaminants to build up on the column. When you run your gradient, the peaks should be about 5X as large, confirming water as the source. Your choices at that point are (a) use better water (buy HPLC grade or make sure your in-house system is working well), or (b) prepare better water chromatographically. We use high pressure mixing systems. In this case, you can put a small C18 column (we like a prep guard column, but a conventional guard will work well), putting it between the pump and mixer. This strips organics and quiets the baseline. If you use low pressure mixing, you can do the same, but must do it off-line, since a column before the pump won't work. Hope these things help.

Could you please give me more details on how to prepare C18 water off line as I have a low pressure mixing system and the HPLC grade water I used is apparently not C18 water(according to the manufacturer)?

A good way to clean up the weak solvent for gradient use in low pressure mixing systems is to wash an Empore Extraction Disk (3M, Varian p/n 1214-5004) w/ a little MeOH or ACN (to pull any hydrophobic nasties out of it first) and filter your weak phase through it. These are 47mm C18 impregnated thick membranes that fit in most standard 47mm filtering set-ups. Things will go slowly compared to 5µ or 0.45µ membranes and you may find that you have to rewash the membrane after every 1-3L of phase, depending on how dirty it is. Amother way is to pull the phase through a C18 sep-pak.

P.S. "C18" water was intended to be read as "C18 treated" - I wouldn't look for it on any HPLC grade water bottles.

I'd be interested in hearing about others' experience with the membrane filter cleanup. We tried one of these from another manufacturer and found that the water was dirtier after we filtered it than before. Any thoughts?

Is it a good idea to wash a column with 100% water after running samples with an acetonitrile:0.02%TFA mobile phase. The samples are run using a gradient that starts with 10% organic and increases to 45%.

No! Don't wash a column with 100% water for most reversed phase apps. All this does is collapse the stationary phase, and it takes a lot of energy (and time) to resolvate it. See a good case study of this in April 99's LC Troubleshooting column in LC/GC. I would simply wash to 100% ACN when you are done -- this will wash out any late-eluters. Then store the column in 100% ACN.

If your system mixes mobile phase under high pressure,another trick is to insert a guard cartridge, containing similar media as the analytical column between the "A" pump and check valve. The spurious peaks stick to the cartridge indefinately. Change the cartridge as needed.

We have had good luck using C18 "extraction disks" to clean up (MilliQ-grade) water for gradient HPLC. We once used cartridges (like SepPak) - these did the job but it would take several hours to process 2-L of water. With a 47-mm disk, we can work at 200-300 mL/min and still get good results. One tip (thanks to SPEC/Ansys) is to make the water about 0.5% in ACN (or MeOH, if that's your organic component) to avoid stationary phase collapse in the extraction disk. This requires a little bit of agebra to calculate exactly what % organic you have in your mobile phase but not too much math.