Chromatography Forum

Evaluating the use of O-phosphoric acid in a MP phase prep. The MP is prep'd @ 60 mM phosphoric then pH adjusted to 4.4 w/ NaOH. Method is isocratic, 1 mL/min, 220 nm w/ 15 min stop. Sx's prp'd in 0.05% sodium metabisulfite.

Problem: The baseline has a very consistent "hump" that starts @ 2 min (shortly after void) and ends @ 12 min w/ the baseline from 0-2 minutes and the baseline from 13-15 min being very stable w/ ~ 0 absorbance.

Any thoughts?

tnx
shawn

Aside from the fact that phosphate isn't a buffer at 4.4 ?

Tom-
Am trying to replace pyrophosphoric acid (pK 5.77) which shouldn't really buffer @ 4.4 either. However, the method is very rugged and provides excellent chrom.. My leader would like for me to stick w/ an inorganic acid for buffering, and none of them have a pK @ 4.4 (HF and tetraboric are closest, but would like to not use them if possible). So I was just evaluating the phosphoric for "fun", and was suprised to see the resultant chrom. Even if the solution is not properly buffered, what would be causing the "hump"?

Shawn,

Do you see the hump only when you are injecting your samples or even with blank runs (injection of water). Does your hump change with different injection volumes?

mtnshawn-

my speculation/concern was that without effective pH control, something in your sample or matrix is smearing out to give the baseline problem.

Kostas' suggestion should shed some light on that question (if you see the same problem with a blank injection)

Blanks (H2O) have humps. Diluent (0.05% sodium metabisulfite) has humps. Humps are also seen on brand new column.

One potential clue, is that during equilibration the baseline is great, so you would think that it is injection related. However, when I perform "fake" injections I don't get the humps. Leading me to believe that the injection pathway (loop, rotor seal, lines leading to the column) is clean.

The needle has been thoroughly cleaned (50 injections @ 50 uL) w/ 0.5% phosphoric in an attempt to get what ever contaminants there may be out of the needle.

Talk about having fun.
sc

From the sparse information, and abreviations I couldn't figure out, it was difficult to think too deeply about this problem, but one comment caught my eye.

What about typical chromatography challenges a buffer pH to any great extent? Unless the sample pH( and we learned nothing about the sample from this question. While, all the available evidence points to the sample as the problem) is buffered greatly different from the mobile phase, it should take but a smidgen of buffer capacity to control pH during elution. A buffer prepared 2, or even 3 pH units from its pKa provides at least a smidgen of buffer capacity. So the original buffer (pKa 5.8 and buffer pH of 4.4) provides lots of buffering capacity unless the sample pH offers a challenge to this capacity.

Why is bisulfite present. Is it a chromophore at 220 nm. Were does it elute?

Bill-
bisulfite is present because this is a modified* USP method for an API (Active pharmaceutical ingredient) and the USP uses the bisulfite to minimize oxidation of the API. We have chosen to keep the metabisulfite in the preparation in order to not deviate to much from the USP method.

Modified*-due to our formulary matrix, we were unable to run the exact USP method and obtain adequate resolution between the API and the preservative. By dropping the pH to 4.4, resolution issues were resolved.

It does have some chromophoric capacity @ 214 nm. It elutes @ the void.

At this point I haven't even evaluated formulary samples. All the work has been done on an analytical standard (purity > 98%) prepared in the 0.05% metabisulfite.

Not sure which abbreviations were causing problems, but here is a list.
MP - mobile phase
sx - sample
sc

I'm stumped! Any hypothesis I had was ruled out by experiments you have already done, especially the water blank injection. It seems like water is displacing something from the coumn, but what? And the problem does not occur with pyrophosphate? Let us know if you figure it out.

What's the column? RP? if RP, which one? Is there anything else in the mobile phase, or just your salt and water?

Have you tried doing an injection of the mobile phase (instead of water or sample solvent)? If so, does the hump still appear?