Removal of Cl ions to prevent adduct formation

[DrPAP]

I have a compound that I am analyzing by ESI-MS using negative ionization. This compound readily ionizes at neutral or slightly alkaline pH. However, the problem is the presence of a Cl adduct. I need to do MS/MS on the [M-H]- ion and not the [M+Cl]- ion because there is a difference in the product spectra (mainly no product spectra with the Cl adduct!). The compound is isolated from biological material using liquid-liquid extraction into a nonpolar phase. Any suggestions on Cl ion removal?

Thanks,
Phil

[Chris Pohl]

DrPAP,

The simplest and most effective way to remove chloride is through the use of a silver form cation exchange sample preparation cartridge. These can be obtained from a number of sources (including Dionex). You'll need to operate under conditions were your analyte is an anionic species in order to avoid losses on the cartridge media.

[MG]

If you are doing LC-MS/MS and the analyte is retained on the column, the problem could be contamination of your mobile phase with Cl-, rather than your sample.

[DrPAP]

I am going to look at the mobile phase contamination (Actually, just organic solvent since this is infusion). I have washed a bottle and cylinder that will be used to store and measure the organic solvents (which is 50:50 chloroform:methanol) with Milli-Q water to assess any improvement from that. The Ag ion-exchange column is a good idea and I'll look into it also.

[MG]

You're infusing it in a methanol / chloroform mixture? That could be your problem. Chlorinated solvents can give you chloride adducts in negative mode. This is reported in

J. Am. Soc. Mass Spectrom. 2000, 11, 932-941

You could evaporate the solvent and reconstitute in plain methanol. If that is too polar to dissolve your compound, you could try mixing in acetone or isopropranol.

[DrPAP]

Thanks MG. I just looked at the paper and was pleasantly suprised to see that the first discussion was using ceramide which are the analytes that I'm working on. The fragment ions of the [M-H]- gives structural information product ions as opposed to the [M+Cl]- which results in only the Cl- ion product. I'm following a paper for quantifying ceramides that uses this solvent as infusion and uses the [M-H]- product ion(s) for neutral loss scanning. I was thinking that I must be contaminating the sample somewhere since I was getting predominately M+Cl and very little M-H during a single-quad scan. The paper's back-extraction of the chloroform stocks using LiOH in the aqeous phase along with addition of small amount of LiOH to the infusion solution shifts to some M-H (probably along the lines of the J ASMS paper with their addition of NH4OH to shift the ratio), but with still predominately M+Cl. Not having a predominance of M-H ions is hurting my limit of detection using the neutral loss scans.

[MG]

Also in the JASMS paper, they mention that some Cl- adducts fragment to yield (M-H)-. I wonder if you could increase your up-front CID voltage (fragmentor, DP, cone voltage, or whatever) to shift the abundance toward (M-H)-, if you haven't already tried that.

[DrPAP]

Good call. I'm using a Thermo Finnigan TSQ Quantum. The Source CID setting does just what you suggest. And yes, I have tried that. I get improvement in absolute response in scan and neutral loss scan with perhaps a small shift in the ratio seen during the scan. Increasing the source CID decreases overall response after a point. It helps, but I still have more M+Cl than M-H. I'm going to see what changing the solvent to a non-chlorinated organic can do for me.