UV Absorbance Difference b/t Buffer A and B WCX Help!

[hkim0327]

Hi Guys,

I am having a few problems with my WCX HPLC system. I am running a WCX-HPLC method to analyze acidic and basic antibody charge variants. The conditions I am using are as follows:

Buffer A: 20mM sodium phosphate pH 7.0
Buffer B: 20mM sodium phosphate pH, 0.5M sodium chloride pH 7.0
Column Temp: 40 degrees celsius
Detection wavelength: 280nm

Gradient: Time Flow Buffer A Buffer B

0.00 1.00 97.0 3.0

30.00 1.00 86.0 14.0

30.10 1.00 0.0 100.0

37.00 1.00 0.0 100.0

37.10 1.00 97.0 3.0

46.00 1.00 97.0 3.0

I am noticing a significant difference between the absorbance values for my buffers A and B when running a Propac WCX-10 column in an Agilent HPLC. Buffer B seems to have a lower UV absorbance, and so I observe a downward baseline drift whenever I run my gradient. I have been making exact calculations when making buffers, so I don't suspect it to be a buffer prep problem.

In order to solve this issue, I made several batches of buffers, but the UV difference has still been persisting. Suspecting it was a system or column problem, I ran the method with and without the column in place, but there was no difference. I even placed solvent lines A and B into the same buffer to see if it was a HPLC line problem, but doing this produced a stable baseline with no problems at all. Moreover, the pump seems to be fine with no salt buildup. I even used a new Propac WCX-10 column to see if the column had gone bad, but there was no difference between the new and old column.

The last thing I tried to solve the issue was to add sodium azide into buffer B so that I could normalize the UV absorbance for both buffers, and I have had some success, but it takes too much volume of sodium azide to normalize the buffers. I have added nearly 0.2% of sodium azide into buffer B to normalize the buffers, and I am pretty uncomfortable with adding this much of azide into the buffers.


Does anyone with a similar situation have any advice?
Is there any other UV absorbing chemical that would help in normalizing the UV absorbance?

Any help would be greatly appreciated! Thank you!

[tom jupille]

Possibly a refractive index effect. UV detectors are somewhat sensitive to RI, especially if the flow cell is a bit out of alignment.

[DJ]

280 nm!?!?

[ScottHorn]

280 nm is pretty standard for protein work.

As far as the original problem, I'd lean towards it being a RI effect as well. As far as matching the absorbance values, I'd use para-aminobenzoic acid. It has such a high absorbance at 280 that just a tiny amount should be all it takes. It shouldn't interfere with your separation. The difficulty will be that since this is (probably) an RI effect as opposed to actual UV absorbance, you probably won't be able to use your spectrophotometer to find the right concentration.

[DJ]

What is the delta in AU here?

If you are hurting for signal this badly, might as well try monitoring low UV (the virtue of using phosphate and NaCL in the first place).

What about a different salt (perchlorate) and some organic in A, B.

[Andy Alpert]

This downward drift in baseline at 280 nm is common with a lot of salt gradients. I too ascribe it to changes in refractive index of the mobile phase. Now, at 280 nm you're picking up the absorption of the tryptophan residues (and of the tyrosines to a much lesser extent). An alternative would be to monitor the fluorescence of the tryptophan residues, using excitation at 280 nm and emission at 320 nm. It's a lot more sensitive than absorbance at 280 nm and is insensitive to refractive index differences. Incidentally, I would say that 9 minutes is too short a time for reequilibration of an ion-exchange column. I'd recommend 20 minutes at a minimum if you want to get reproducible retention times for the least well-retained analytes.