[Nescafe]

Hello,

I was wondering what is the best method (NMR, UPLC) if you know you have a mix of two diastereomers and nothing else to determine dr. Also, you know that let us say S diastereomer is the major.

Thank you in advance,

Nescafe.

[discodermolide]

The most accurate method would be HPLC assuming you have a good separation.
NMR can also be used but you may need to experiment with the best solvent, d6-benzene may cause an aromatic induced shift, it may or may not be in the direction of signal separation, you need to experiment.

[Nescafe]

Good to know.

So by HPLC would be to measure UV AU, what if the compound is not UV active, could you use a LCMS and use the ES+ trace or is that not as accurate?

[discodermolide]

LCMS should work. The MS is only a detector. Crucial is the LC separation of the diastereoisomers.
Can you separate them on a TLC plate?

[Nescafe]

I have not had much luck separating them on the TLC, although I am sure it is possible if I tried more gradients. Fortunately, I do see a separation on the LC which goes something like this:

<Peak 1, retention time 0.38 (5% area) v.s. Peak 2, retention time 0.59 (95%)> Both by MS have the same mass of 191

Only concern is that I then made the diastereomer of Peak 2 and was expecting that by LC I would see the reverse scenario, but instead I see this:

<Peak 1, retention time 0.27 (99% area) v.s. Peak 2, retention time 0.67 (1%)> Both by MS have the same mass of 191>

Why are the retention times different, it is the same condition and I did my best that the concentration at 4uM.

Thanks,

Nescafe.

[discodermolide]

This must be something slightly different in the instrument. Flow rate, detection, concentration, solvent etc.
I would say that both results are probably ok.

[Dan]

Why are the retention times different, it is the same condition and I did my best that the concentration at 4uM.

Try running the samples back to back. I have had some small issues with drifting retention times (though I think both peaks drifted in the same direction, but it was a long time ago) in chiral HPLC where the peaks in an enantioenriched sample did not match fantastically well with the racemate prepared several months previously. After digging the racemate out of the freezer and running the two samples consecutively the retention times matched.

Presumably NMR indicates that the two samples contain the same compounds in different ratios? At 99:1 it's going to be pretty difficult to make sure the minor component is what you think it is. How many isomers can you reasonably expect given the structure of the molecules and the methods of preparation?

[discodermolide]

The best way to check is to take the second sample you made and spike it with some of the first and re-run the HPLC. Then you can be sure you are dealing with the same compounds.

[Nescafe]

I actually did spike it. So when I ran the second sample I did not even see anything other than one peak at 0.27 so then I spiked it with the first sample, which according to the same LCMS machine had a retention time of 0.59, but when mixed it shows at 0.67.

So just to clarify, my second sample, call it sample B, which is the diastereomer of sample A, only had a peak at 0.27. I spiked sample B with A and then saw two peaks but the retention times do not match.

I can try running them one after the other.

[discodermolide]

Did you manage to check the NMRs of all the samples? I am assuming you can assign all the signals.

[Nescafe]

Yeah the NMR checks out and the peaks are all assigned.

The NMR is as follows

First sample

Distinguishing Large peak at 2.3ppm (1.0 integration)   Small neighbouring peak (0.05 integration) at 2.1ppm

Second sample

Distinguishing Small peak at 2.15ppm  (integration 0.01)   Large neighboring peak (1.0 integration) at 2.0ppm

Just when I bumped into this issue of retention time difference I started second guessing. I mean, it is interesting how individually they have different retention times than when mixed together.