Reproducibility problems with ESI

[sahram]

Hello,

first of all I am a beginner with electrospray. Now I am using LC/MS with electrospray ionization in negative ion mode. I try to quantify some acidic carbohydrates with 0-3 sulfate groups/molecule.

First of all I don't want sulfates to drop off so I have tried to find mild conditions and still get some ionization. Addition to sulfate loss the molecules also tends to fragment readily.

I have used Agilent MSD low resolution detector. Eluent is 50mM Ammonia plus 20mM Ammoniumbicarbonate.

The main problem is that ionization seems to differ from day to day. The peak with most intensity can be 1)double charged, 2)singly charged with one sulfate loss. All the parameters are the same. I think that when I have higher concentration of analyte, the 2) comes predominant. I also get molecule with one charge and three charge. It would be ideal to ge singly charged only.

I have to smooth my raw data, because on low concentrations, chromatograms look awful, because ionization is so low. My lowest concentration is now 450pmol/ml but of course the column makes it more dilute. Flow rate is 0,5 ml/min and peak is 1 minute broad.

Please, can You tell me what to change to
1) get better ionization without loss of sulfates or further fragmentation
2) to get same ionization pattern every day, preferentially singly charged
3) to get same ionization intensity every day
4) what else I should think

Thank you,
sahram

[MG]

1) Lower fragmentor voltage. If you are as low as you can go without losing parent ion signal, try lowering source temperature. Also, particularly with negative ion mode, make sure your electrospray voltage (I think called VCap on that machine) isn't too high. If you see a corona discharge from your ESI needle (turn off the lights in your lab to make it easier to see), then your voltage is definitely too high.

2) See if lowering the mobile phase pH helps. Also make sure your compounds are well retained on the column, so that injection solvent doesn't influence the ionization.

3) The above tips will help, but there are no guarantees. It is common to have to run calibration standards daily and check standards (or QC's) every ten samples with LC/MS, because response can drift. One-thrid of developing a rugged LC/MS method is in using good chromatographic techniques (which you can learn about by hanging around this board); another third is in properly optimizing your MS for the analytes and LC conditions in use. The last third is sample-prep.