EPA Method 544 Microcystins

[James_Ball]

Has anyone here ran this method? LCMSMS with MRM.

We are currently developing for this method and I have most of it working except where it comes to the final extract composition.

The method says to bring the final extracts up in 10%water/90%methanol. The initial mobile phase is 90%water/10%methanol. When I inject the standards in this solvent I get a spike at the beginning where the solvent front comes through with hits for all the analytes, then when the analytes elute the peak shapes are not great and have low sensitivity. The method uses a solvent divert at the beginning so they have no example chromatograms for this region of the run, since I am not using the divert I do see it.

If I bring my extracts up in 50/50 water methanol this goes away and the peak shapes are excellent. The method states you can't alter any of the extraction steps but you can dilute the samples, but if I dilute to change the composition then I may not be able to achieve the detection limits required. I have never had any luck injecting high levels of methanol into a mostly aqueous mobile phase, is there any trick to making this work, or is the method simply poorly written? Even if I begin with 70%water/30%methanol as the mobile phase I get the same problem with the analytes running through the column with the solvent front.

[Rndirk]

It's weird they propose to inject high organic in a high water composition since this issue is well known...

Is (partial) solvent change instead of dilution an option? It's cumbersome i know...

Another, more drastic change would be to switch to HILIC separation, where injection of high organic is not a problem.

[Jake]

Agreed - not the best approach for quality chromatography.

One way to improve chormatography with these conditions (high organic sample solution on high aqueous mobile phase) is to reduce injection volume relative to column volume. The other is a rapid B-solvent ramp. Not sure if you can change these with the restrictions of the methord, or if they're too obvious, but that's all I've got.

[James_Ball]

I have been working with C18 phases so far, the method was written using C8 Kinetics column 2.1x100mm 2.6um. My columns are Waters Xselect HSS T3 2.1x150mm 3.5um and Restek Raptor ARC-18 3.0x100mm 2.7um. We are probably going to get the Kinetics C8 and try it, but would it give any advantage over the C18 columns in this situation?

I may try to do the 50% sample solution and see if we can get it approved. It works great, but locking that down as something that can't be changed could be a problem. Unless when it says "can not alter the concentration step", it can be argued that the final solution is after the concentration.

[Jake]

You mean Kinetex by Phenomenex, right? In our experience with several classes of analytes and C8 and C18 Kinetex columns, they'll work much better, worse, or comparable to other C8 and C18 columns. We've never been able to predict and end up just trying our various C8 and C18 columns.

[Darby Gloss]

hi, I am a new here, nice to see u, and do you want you buy produces with high quality and inexpensive, there are a lot of usb watch +8gb -spy, you can see and I believe you can find a good one.

[Rndirk]

In theory I would say you'd be best off with a column that gives the most retention during the initial mobile phase composition for the analytes that elute first, since this would lead to focusing. A higher ramp (as Jake says) could also lead to focusing.

I agree that lowering the injection volume would be better, but you have the same problems as with dilution (detection limits), it's basically the same thing.

[James_Ball]

Trial and error is usually the way I end up doing it also. With so many variations of silica for each stationary phase you really can't predict what will happen, unless you are someone who works every day with each type.

If I had more sensitivity I could reduce the injection volume until the problem was mostly negated, but I was seeing it even with 10ul injections at higher concentrations. Of course it is somewhat of a catch 22 situation in that if I can get certified for the method then we will have funds to upgrade the instrument, but we really need to upgrade the instrument to get certified.

I have also found that altering the mobile phase from 20mM Ammonium Formate as A and Methanol as B to 5mM Ammonium Formate + 0.1% Formic Acid as A and 0.1% Formic Acid in Methanol as B improves both peak shapes and sensitivity. There are also several analytes that end up doubly charged, which is listed in the method but I have one extra, the surrogate, that is doubly charged and I can't get a singly charged ion to form as they do in the method, at least not at any decent abundance.

[James_Ball]

I am looking more closely at the method as written and it shows that they are doing a 10ul partial loop injection into a 20ul loop. If you are getting mixing in the loop, would that not be essentially a 1:1 dilution of the 10/90 sample solvent with the 90/10 mobile phase, which would result in the 50/50 injection solvent I have found gives good peak shapes?

[James_Ball]

I am thinking up too many things to try on this one.

Would it work possibly to put an inline static mixer after the injector to mix the sample and mobile phase before it gets to the column, then allow the column to refocus the analytes? I am thinking along the lines of one of the 5-20ul inline mixers for UPLC mobile phase mixing.