Chromatography Forum

I struggled with a suspected contamination problem with an Agilent HPLC 1200 system. I tried to eliminate the problem for 2 weeks but in vain. I desparately need some expert advice and help here.

I used a C18 column to analyze a reaction mix consisting of N-acetylcysteine (NAC) and nitrite in HCl. S-nitroso-N-acetylcysteine is the derivatized product to be measured. But since I used an excess of derivatizing agent (NAC), the latter is present in about 1361-1461ppm in the mix.

1. Mobile phase used: 15% ACN in water with phosphate buffer.
2. isocratic
3. At 210nm, NAC peak appears at solvent front but the UV spectrum is non-specific.
No sign of contamination (of the non-specific UV spectrum) in C18 column even during wash with gradient of 5% ACN to 70%ACN.

The contamination problem occurred after I changed to HILIC column and ran some samples.

HILIC condition as follows:
1. MP A: 0.1% formic acid in water; MP B: 100% ACN
2.
Time (min) A% B% Flow (mL/min)
0 5 95 1
4 95 5 1
7 95 5 1
17 5 95 1
18 5 95 0

3. In blank sample after sample, I see a peak and a plateau of UV absorbance at 210nm.

I washed the whole system (including column) with different ACN concentrations, then 0.1% formic acid in IPA: water (1:1). Contamination still present, though halved in height after wash.
I disconnected the column. Then washed with 3M HNO3. Contaminant still there.
I tried no injection, the plateau is still there. I connected the purge valve directly to DAD, the plateau is still there.
I can’t figure out the problem.
Please help. Thanks in advance.

I would guess that the plateau is simply UV absorbance mismatch. Formic acid has some absorbance at 210 nm. There's not much you can do about it short of adding just enough formic acid to the ACN to match the absorbances (which is difficult to do and is usually more trouble than it's worth).

The contaminant is more puzzling. As I read your post, it showed up as a peak even with no column connected, which raises a question about whether what you were seeing then was the same as the "contaminant" you saw with the column connected. Is there a chance that the contaminant is gone and what you're seeing is "t0 noise" from the injection pulse?

Formic acid has some absorbance at 210 nm

I checked the UV spectrum of formic acid. It looks exactly the same as my "contaminant".

As I read your post, it showed up as a peak even with no column connected,

Sorry, I did not mean it showed up as a peak with no column connected. When no column was connected, it showed up as a plateau.

I am very grateful to your suggestion. I think my system has not been contaminated. The "contaminant" is actually formic acid. I should use other buffer instead of formic acid in the mobile phase A. One further question: Is there any recommended buffer for HILIC column and UV detection?

A million thanks!

I wonder if high conc. of acetonitrile also give high absorbance at 210nm? Is methanol a better mobile phase for UV detection?

I wonder if high conc. of acetonitrile also give high absorbance at 210nm? Is methanol a better mobile phase for UV detection?

Acetonitrile is UV transparent down to 200 nm but methanol has a very high absorbance at 200 nm so it would not be a good choice for your analysis.

Is there any recommended buffer for HILIC column and UV detection?

My understanding of the HILIC mechanism is that the identity of the buffer should not have much effect on the chromatography. Phosphate is the usual choice for low-UV so long as you are not thinking of ever moving the method to LC-MS.

Thanks. I'll go for phosphate buffer in ACN.

Phosphoric acid is very soluble in pure ACN but as soon as you add ANY base (i.e. make a salt), the phosphate precipitates.

Phosphate has very limited solubility at ACN/H2O concentrations >80 v/v% and the solubility depends on the pH, i.e. the valence of the ions.

Below is a good LCGC article giving approximate solubilities of different phosphate salts at different pH's in varying v/v% ACN/H2O.

http://chromatographyonline.findanalyti ... rticle.pdf

Carls, your reference is a good one. Thanks a lot!