RT shift observed during chiral analysis

[ntruong]

Hi All,
We have one validated LC method (isocratic) to determine enantiomeric purity using mobile phase system 97:3 Hexanes: t-butyl alcohol with Chiral AD-H column. The main peak initially comes out at 65 minutes and then gradually drifts past 100 minutes. The run program intially set at 80 minutes, now it is extended to 120+ minutes (sometimes 150 minutes) in order to capture the the main peak (I have total 5 "pricey" chiral columns and they all behave the same). I tried them all on 4 Agilent 1100 series and got the same issue (I also allow at least 12 hours of equilibration). I also observed pressure fluctuation on 4 LCs. As part of investigation, which I strongly believe mobile phase is the culprit, I prepare another mobile phase system 95:5 Hexanes:t-butyl alcohol and it works like a champ. I set up sequence to run with this new mobile phase for 24 hours and the Retention time of the main peak (comes out at 38 minutes)doesn't shift at all (I have tried on 3 LCs and I don't see any pressure fluctuation). Can anyone explain to me why? It seems to me I got unstable mobile phase using less organic modifier (t-butyl alcohol at 3%) compared to 5% t-butyl alcohol. I'm new to this chiral technology and I
do need your help.
Thanks in advance,

[tom jupille]

I suspect it may have less to do with "chiral" and more with normal-phase separations on cellulose. Trace water can play an important role in such separations. I wouldn't be surprised to find that retention decreases on your 3% BuOH mobile phase if you saturate it with water.

[Bruce Hamilton]

That's an improbably long retention time, especially as Diacel says that with methods longer than 20 mins, users should consider another column. I believe the AD-H is an amylose column, and I've only experience with OD-H ( cellulose ), but should be similar.

Given that the long retention, some variables will have more effects.

Mobile phase water content as Tom suggested above, use Karl Fischer to check the two batches.
Column temperature, assuming it's not too high I'd increase it to see the effect on retention time.
Sample solvent should be same as mobile phase
Instrument preconditioning, ensure that your storage and preconditioning are exactly as ( or more protracted ) specified in the column notes, especially ensuring that the column and instrument is really well flushed after use, as Diacel emphasize that for OD-H.

We get really good repeatability ( at 17 min retention ) if we spend time flushing and preconditioning using ethanol exactly as specified in our column notes.

Personally, I'd revisit the method to get retention closer to 20 minutes, even if that means a different column, column temperature, or mobile phase, life's too short in the 21st century for methods longer than 60 minutes

Bruce Hamilton

[JM]

R U using a premixed mobile phase or two chennels??

JM

[ntruong]

JM:
The mobile phase is premixed into one reservoir, isocratic, one channel.
BH:
When I changed the mobile phase to 95:5 and the problem is solved. No more retention time shift as well as pressure fluctuation (22 bar - 38 bar). Note: my boss said it was the check valve and it has nothing to do with mobile phase. The main peak is at 38 minutes. We will partially re-validate the method (Robustness study only). I still don't know why 95:5 (hexanes: t-butyl alcohol) is more stable than 97:3 system.
TJ:
Is it possible that the right amount of alcohol modifier in the mix might do the trick such as equilibrium? I didn't get a chance to check for the water content but I don't think there is any water in the alcohol though.