separation of positional isomers

[Mike H.]

Quick question. Other than using a phenyl column, what are my options to resolve positional isomers?

[HPLC chemist]

Depends on the isomers of what? In most cases in the past the answer was GC.

[tom jupille]

. . . or porous graphite, or C18 / C8 with low surface coverage. As HPLC Chemist said "Depends on the isomers of what"

[mattmullaney]

Hi Mike H.,

E/Z isomers for example, may be easily separated using (RP) HPLC with polar-embedded phases such as Agilent Bonus RP, Thermo Polar Advantage II or Supelco RP-Amide.

[Mike H.]

I would think these would be simple isomers. It's hard to give structures without knowing what is and what is not proprietary. I'll try. Visualize a benzene ring with identical moieties at position 1 and a single CN group at the para and meta positions.

I've read that these types of isomers work well with the adsorption mechanism. Other than some chiral methods that utilize normal phase, I have no experience with normal phase. What column would I need if I wanted to try this? Is this the easiest path?

[mattmullaney]

Hi Mike,

Okay then, so these are ortho, meta and para isomers. Looks like you read a bit of "Introduction to Modern Liquid Chromatography", which is a good start. Yes, Normal phase chromatography would work best from first principles, and you should be careful as newer HPLC do not always have friendly materials to normal phase solvents as installed. Generally LC running NP methods are dedicated these days in my experience. Modifying the LCs to run NP isn't hard, but it will incur some cost.

Is there any TLC separation(s) for these compounds? Will provide a hint toward the eluent composition and phase (most likely bare silica as opposed to bonded NP material).

Is there any literature describing the separation of some or all of these analytes? Sometimes minutes doing research can save hours in the lab. I like sciencedirect myself, but sometimes even Google can help with this.

If the LC is NP ready, there is no reason for concern; it's just a different way to perform a separation--note Dolan's caution to "half-saturate" the NP eluent with water and you'll be pleased, I think.

[Mike H.]

What does half-saturate the normal phase eluent with water mean?

[mattmullaney]

Hi Mike H.,

If you happen to have "Introduction to Modern Liquid Chromatography", refer to Section 8.5, pgs. 392-394. In particular if you use a bare silica NP column, retention times for peaks can vary when the eluent (mixture) contains different water content(s) from batch-to-batch or from uptake if the lab humidity is very high. To ameliorate this, prepare the NP eluent, divide the prepped NP eluent in half and place in a large separatory funnel and place the other half into the eluent reservoir. Add a (small) quantity of water to the separatory funnel [You can get an idea in advance of how much water to add by using a table of water solubility values in the solvents that comprise the NP eluent.] Shake the sep funnel to add water to that portion of the NP eluen--this affords water-saturated NP eluent...then add the water-saturated NP eluent to the relatively-drier NP eluent in the separatory funnel and mix well.

You now have half-water-saturated NP eluent!

[Mike H.]

I was able to separate one pair of isomers with a Phenomenex Biphenyl column.
Conditions were 40°, 50:50 0.1% HCOOH:MeOH, but I'm now experiencing some bad tailing ~2.5 USP. The isomers are small molecules containing 2 aromatic rings with a carboxylic acid group, a nitrile group and a CF3 group. Can anyone give a best guess on what is causing the tailing?

Incidentally 0.1%HCOOH:ACN does not separate the isomers.

[mattmullaney]

Hi Mike H.,

Can you describe the Phenomenex Kinetix stationary phase a bit? I'm having trouble finding a description.

Inject 1/2 the normal volume, see if tails are maintained, Is phase endcapped? If so how? (OK--later I found out that the Kinetix Biphenyl phase is encapped, I'll think further about tailing. One tailing cause is phase overload, which is why I suggest cutting the inj. vol..)

[lmh]

ACN wouldn't be a good choice on the biphenyl column because it uses pi-pi interactions between the pi electrons in the benzene rings of your analyte and the phenyl groups on the column. ACN has so many pi electrons it out-competes. If you use ACN with phenolic columns basically they become a not-very-retentive reverse phase column.
Not sure about the tailing; possibly you're too close to the pKa of your acid group??

[lmh]

... should have asked: I assume you've tried the obvious, straightforward reverse phase?

[Mike H.]

Overload didn't seem to be the problem, since a lower injection volume didn't affect the tailing. What did help was lowering the temp from 60° to 40°. When doing this tailing dropped from 2.4 to 1.3. The upper 1/3 of each peak broke towards the x-axis. In other words the slope of the upper 1/3 of each peak is lower than the slope of the lower 2/3 of each peak. I'm hoping this is an equilibration issue.

[mattmullaney]

Hi Mike H.,

My thanks for getting back--and congratulations on figuring out what was going on. Along the lines of Imh's suggestion above, perhaps moving away from ACN to a blend of ACN/MeOH would be helpful in further reducing tailing?

I went back to my own experiments throughout my (too many!) positions, to date I've used a biphenyl phase when that phase was fairly new..and I re-learned that I used an aqueous buffer in a gradient with 60:40 (v/v) ACN/MeOH.

[Mike H.]

Update: The second injection at 40° again showed the tailing. I found a miracle additive via google (0.1% TFA) which solved the tailing and improved the separation.