LC-MS/MS pH=10.2 Xbridge, m/z=141 contamination

[Jakob Jornil]

Hi,

I have some problems with contamination on my Waters Micro LC-MS-MS system, I hope somebody out there have some ideas to help me!

HPLC method
Solvent A: 5mM ammoniumformate in H2O adjustet to pH=10.2 with ammoniahydroxide, 5% acetonitril.
Solvent B: Acetonitrile.
A gradient program is run. Column: Waters Xbridge.

I am analyzing amitriptyline and metabolites, there is a nice separation because of the basic pH.

Initially I used a Phenomenex Gemini column, but they allways died after ~500 injections with peak tailing and rising backpressure. Switching to Waters Xbridge column have so far given me 1000 injections and there is still nice peaks.

The problem: In MS-SCAN mode I have a huge contamination at m/z=141 - and also a peak at m/z=59. Running pure metanol into the MS from a syringe the m/z=141 peak is still there. Cleaning ths MS (Cone, extrator cone and hexapole/quadrupole) helped the situation. But when I used my HPLC method again the contamination got back. Where does this conatamination originate from? (and what is it?)

The most obvious reason is breakdown products from the column, origination from the use of the basic solvent A. I have tried different basic solvents but the described solvent is way the best for my analytes. Ammonium bicarbonate was not good, giving irreproducible results. Other suggestions for a basic solvent for LC-MS - would also be nice!

If you have experience running LC-MS at basic pH you are more than welcome to share your knowledge (column lifetime, solvent selection and other problems)!

Best regards Jakob Jornil

[JMB]

Jakob,

I suspect that you may be looking at adduct ions of the type,

[(H2O)2]Na+.................m/z 59

{[(MeCN)(H2O)2]Na+........m/z 100.....is this also present ??}

[(MeCN)2(H2O)2]Na+......m/z 141.

Na+ derives from the glassware.

Does infusion of MeOH completely remove the m/z 59/141 ions ??

Do you see the equivalent adducts at m/z 91 and 123 formed with MeOH ??

You can confirm (or deny) this explanation by generating the MS/MS spectra of 59/141. Alternatively, adjust cone voltages to fragment them ?

Let us know what you find.

JMB

[Jakob Jornil]

Hi JMB,

Thanks for the reply! It sent me to the MS looking at fragments ions. Here is what I found:

At a cone voltage of 5V, a eluent of acetonitrile/water and doing MS-scan I found ions at m/z: 59, 100 and 141.
Fragmenting (MS-MS) m/z 141 gave ions at m/z 100
Fragmenting m/z 100 gave ions at m/z 59
This seems to confirm your proposed adductions.

Adjusting the cone voltage to 25V gave an increase in m/z=141 response. Fragmenting the 141 ion (MS-MS) at cone volatage of 25V gave a different daughter ion pattern with ions found at m/z: 42, 58, 71, 85 and 112. So these fragmentions seems to originate from another m/z=141 mother ion, which I guess is my contamination ion.

Since I posted my original message the contamination at m/z 141 in the MS-MS has been reduced by a factor 100. Running pure acetonitrile through the system over night might have solved my problem. But I still need to figure out what is happening since the contamination sometimes reappears. One incidence where the API-gas was shut down by accident, and the LC still pumping solvent made the 141-contamination reappear.

Using the syringe infusion with MeOH doing a MS-scan gave the most intense peaks at m/z: 50, 59, 64, 87, 96 and 105 (not 91 and 123), also at conevoltage of 5V.

Does this make any sense to you?

With the best regards Jakob