Deleted

[TimB]

Deleted

[JGK]

Most materials will absorb @ 190 nm


the "mess" in front of the peak is probably the solvent front and un-retained extract components.


When you talk about blanks, are you referring to "blank" extracts or a non- extracted blank?

[Bruce Hamilton]

Most materials will absorb @ 190 nm

I agree, including 0.1% formic acid. I would suggest that very little energy is passing through, and refraction and other effects ( including stray light ) may modify the signal - the negative peak is probably very real, as your mobile phase probably strongly absorbs at 190nm.

I would review the the mobile phase absorption, and perhaps select a slighly-higher wavelength and/or reduce formic acid concentration..

Bruce Hamilton

[Alex Buske]

Hi,

Your t0 is 5.5min, at this point much of stuff (including solvents) injected eludes. Your analyte comes at 8 min, thats an k' <1. Try to increase the Flow rate (1.3 ml/min is appropriate for 250x4.6mm columns) and to decrease the organic part. That should shift your analyte out of the "injection junk".

Alex

[Vlad Orlovsky]

Actually K' is 0 for flow rate of 0.3 ml/min on this column. Your compound is not retained on the column and elutes in void as well as other "junk". Void is APPROX. 7.8-8.3 minutes (and not 5.5 min) on 4.6x250 mm column at 0.3 ml/min. At 1 ml/min void is 2.4-2.6 minutes on regular 4.6x250 mm, 5 um, 100-120A column. You are running more than 3 times slower rate.
You need to change conditions or column. Try to run 0% organic with 0.1% phosphoric acid.
Is your compound ionizable or not?

[TimB]

Wow thank you for all the quick responses. I should mention that I am not the developer on this method, and it has been "inherited." Upon further research, I found it was taken directly from a paper that uses this method with LC-MS. There was really no further development.

Based on your suggestions, I am going to try lowering the formic acid concentration, increasing the flow rate, using 0% organic and 0.1% phosphoric acid.

JGK- when you asked about the blank, the extraction diluent is water, and I just run a sample of millipore water as the blank (same also being used for extraction).

I believe the compound is ionizable since the method was initally for LC-MS.

[Vlad Orlovsky]

UV cut-off for formic acid is 230 nm, the reason they used formic acid is detection technique. Don't even bother running it with formic acid, go with UV transparent mobile phase (water+inorganic acid)

[HW Mueller]

TimB, ionization in the MS does not necessarily mean that ionization occurs in the aqueous phase. I think Vlad just wanted to know whether you really need the acid in the mobile phase.

Incidentally, that original method seems to be an example of what I mean with "sloppy LC" in conjunction with MS.

[mohan_2008]

0.3 ml/min is not enough flow rate to equilibrate a 250 mm column - may require extensive equilibration times specially at high aq. ocncentrations.

There is an equilibration issue in your method.

[Vlad Orlovsky]

and what exactly you need to equilibrate in isocratic run? Any rate is enough to equlibrate, just depends how long (3.3 times longer than at 1 ml/min)?

[lmh]

- a mini-rant agreeing with HWMueller about LC-MS operatives who publish inadequate methods:

HWMueller, you're right, it's an extreme case of sloppy LC-MS. Where, in the past, I've suggested LC-MS is easier than LC on its own, I wouldn't for a moment condone running a method where the column really isn't doing anything.

There's nothing sloppy about using formic acid if you intend to use LC-MS. It often gives much better signal in MS than even volatile buffers such as ammonium formate; formic acid isn't a buffer, but the bottom-line test is what the mobile phase additives do to the peak shape/retention time in the context of the sample matrix, and whether they achieve constant values from run to run.

On the other hand, merely reducing the flow rate to 0.3mL/min, presumably to keep the mass spec happy, rather than simultaneously reducing the column diameter to match the new flow rate is plain wrong. If you must use a huge column, keep the flow right for the column and split the flow to the MS.

Similarly, running a solvent system where the analyte is not retained is fairly pointless; it's actually worse in the context of LC-MS because it's a recipe for cosuppression by all the other things that will appear at the injection front, not to mention changed ionisation efficiency due to residual injection solvent.

I wish journal reviewers and editors would be a little tougher on methods that look bad. They merely waste other people's time when it comes to reproducing the work elsewhere.

rant over; you lot are tolerant folks!

[HW Mueller]

In this connection I am curious as to whether TimB gets a negative peak when injecting pure Water (blank), and what RT it has.

[TimB]

It has been so busy here that I have not had any time to go back and look at the method or any of the chromatograms (It's sort of low priority now). I hope to work on this method starting next week sometime.

Edit: Until then, thank you for all of the suggestions!

[TimB]

So it's been a while since I started this project. But we haven't had a demand for this analyis in months, and i've had to do other things. We recently had a request to do this, so my boss gave me the OK to work on it again.

Here is the new chromatogram:


Mobile phase is 0.1% phosphoric acid in water.
Inj vol is 10 uL
Flow changed to 1.5 mL/min.
Isocratic run for 16min.
Detection: UV @ 190 nm
Column: Waters Atlantis C18 250x4.6mm (5um particle size).


The first two peaks are unreacted species in the standard (which we expect), and the third is our analyte. The chromatography could probably even be improved more, but the peak shape is acceptable for us. My boss was very impressed. Thank you for the help chromforum.

Anyone have any further suggestions for improving? Gradient with an organic? column heating?

[Bryan Evans]

Looks like you made progress. Is your sample dissolved in mobile phase (0.1% H3PO4)?