False Positive Integrations for GHB

[Lunar]

Hello,

I am developing a method using LC-MS/MS to confirm GHB, 1,4-BD, and GBL. I have successfully created profiles for each compound except the retention times. Every time I run to get the retention time, 1,4-BD at higher concentrations always integrates on the GHB method. In addition, GBL is appearing at GHB and 1,4-BD as well. As such, I cannot proceed until I better differentiate the two.

Currently, my settings include a flow rate of .2 ml/min, ES+ mode, a gradient of mobile phase A at 100% at 0 min, 90% at 2 minutes, and 20% at 3 min with a total flow time of 5 minutes. Mobile phase A is water with .1% ammonium formate, and mobile phase B is 0.1% acidic (formic acid) methanol. I am running on a HPLC-MS/MS with a C18 column.

The current retention times I am getting are: GHB at .753 min, GHB-D6 (internal standard) at .77 min, 1,4-BD at .77 min, and GBL is variable.

If anyone knows anything I can do to better differentiate the two I am more than willing to try it. I have changed the flow rates, the gradients, and total flow time but to no avail.

Thanks

[TylerSmith123]

Hi Lunar,

Could you provide us with more information about your column (ie size)?

I'm not sure what the acronym are, and I assume 1,4-BD is 1,4-benzodioxane but I could be completely wrong. Anyway, I have a few question for you regarding your separation and sample prep. When you prepare samples what is the matrix? Is it the 100% water from your starting conditions or something different? Additionally, what's the point of the acid coming in at all? According to you two of your compounds have eluted prior to seeing any acid and it's only mixed into your b-solvent which I find quite peculiar? Typically, you'd make your buffer and then adjust the pH, but now with your buffer and MP B mixing there is no telling what is happening to your mobile-phases pH. Additionally (and I assume this will be small) what is the dead-volume? I can tell you're using a small column, I'm just curious.

[Lunar]

Yes I may!

We are using a Acquity VEH UPLC C18 1.7 um column (sorry I know I said we used an HPLC I meant a UPLC my bad).

The method we are using to acquire the retention times is as followed:
- First take the standards and evaporate until dryness
- Next, re-constituting with acetonitrile water 50:50 solution (I attempted dH2O and it did not come up with anything for some reason).
- Afterwards, inject them directly into the LC cups
- Infuse the machine with the standard. Buy this we place our sample in one of the reservoirs, change the settings from LC or Waste into Infuse after purging the system. Then we go to Intellistart, added our compound to the list (molecular weight, parent M/Z predicted point, etc.). In positive H+M mode we run with our tune to acquire the cone voltages and collision times.
- When completed with all samples, we added the data into out MS method and our processing method excluding the predicted retention time, that was kept at zero with a +-3 minutes to scan the entire spectra for predicted times.

We are using acidic methanol and water because GHB is a weak acid and will convert to GBL if kept at neutral or basic conditions. Quite frankly, I don't know why it's fine to use acidic conditions with GBL it is simply just what the procedure we are basing off our method off of did (I can attach a citation to the paper if you'd like). 1,4-BD stands for 1,4-Buitendiol because with GBL it will naturally convert to GHB invitro within a minute or two.

Thanks so much for being willing to help!

[TylerSmith123]

Hi Lunar,

What I'm thinking right now, depending on the length of time these compounds are in your injection solvent, GBL may be variably protonated and de-protonated. So while sometimes, upon injection, you'll see split peaks, or two peaks for the protonated and unprotonated versions. I'd attempt the method with acid in your MPA, as well as changing your injection solvent to include methanol rather than ACN and some acid-- either TFA or Formic for MS. In that short amount of time I doubt the MPB acidity alone will guarantee protonation. This may be what's causing your concentration issues because of the conversion, inadvertently giving you more of one product than the other.

I'm curious if your separation can be done isocratically as well rather than this short gradient-- which may be a reason for some of your inconsistencies, but I'd say that would be the last suspect unless you see other symptoms.

Do you also have a problem with the standard eluting right on top of one of your peaks-- this could also contribute to the peak's area expansion into your other analyte. Is it possible to use a different internal standard?

The last thing I would mention would be to maybe tweak around with the integration settings? This would probably impact a few things you already have programmed, but maybe it will fix them integrating together?

Sorry I couldn't be more of assistance,
Tyler

[vmu]

sorry I know I said we used an HPLC I meant a UPLC my bad

No problem. UPLC is just a variant of HPLC.

[Lunar]

Hello Tyler,

I cannot do it isocratically

I also am now swapping out the reconstitution solution to methanol. After that I will look at the data and then add an acid to MPA.

In addition, currently I do not there is another internal standard for GHB. I have reached out to a neighboring lab because they have made method validations before and they have suggested to have a ratio tolerance because the areas and ratios of the false integrations are extremely small as well.

Thank you so much for your time and knowledge! I very much appreciate it!

[James_Ball]

Hello Tyler,

I cannot do it isocratically

I also am now swapping out the reconstitution solution to methanol. After that I will look at the data and then add an acid to MPA.

In addition, currently I do not there is another internal standard for GHB. I have reached out to a neighboring lab because they have made method validations before and they have suggested to have a ratio tolerance because the areas and ratios of the false integrations are extremely small as well.

Thank you so much for your time and knowledge! I very much appreciate it!

As mentioned above I would use Water with 0.1% Formic acid in MPA, this would keep the acid concentration the same through out the gradient as it is what is in the Methanol in MPB.

Have you injected each analyte through the column separately and found the problems or just when they are all in the same mixture?

There should not be a problem if the internal standard is eluting at the same time as the target, the isotopic labeling should offer the mass separation needed for identification of each in a co-eluting pair.

[Lunar]

Hello Tyler,

As mentioned above I would use Water with 0.1% Formic acid in MPA, this would keep the acid concentration the same throughout the gradient as it is what is in the Methanol in MPB.

Have you injected each analyte through the column separately and found the problems or just when they are all in the same mixture?
There should not be a problem if the internal standard is eluting at the same time as the target, the isotopic labeling should offer the mass separation needed for the identification of each in a co-eluting pair.

So I did swap out the mobile phase. In addition, I ran batches comparing various reconstitution solutions, the one that had the best and most differentiable peaks was acidic water.

Now I run into another dilemma. Either I did not add enough GHB-D6 to the batch or I very much have no idea where the peak is. It is not integrating and, based on many papers, it looks like it should have a similar retention rate to that of GHB. Currently, we are running it for 10 min total because D6 did not integrate at 0-4 minutes and we were hoping to see it later on. I will, later next week, go through 2 minute intervals on the MS method to see if then I can identify GHB-D6. I will keep reply once I have that data as well.

Thank you for your help!