Chromatography Forum

Hello,

I have a separation problem; my two components can be separated by a Hypersil MOS-1, but the resolution is about 1.4 to 1.5.

One of the optimization strategies will be the screening of other non-endcapped C8-columns. Please give me some hints (brands, manufacturers).

Thank you very much.

Florian

http://www.usp.org/USPNF/columns.html You can go to the USP column selectivity database (http://www.usp.org/USPNF/columns.html ) and look for either "most similar" or "most different" columns.

Dear Tom,

I was already in the USP database, but the number of C8 columns is limited therein. Especially, the Hypersil MOS-1 is not outlined.

My aim is to get a series of proposals because my first search on the manufacturers webpages yields primarily endcapped C8-phases

Florian

There are over 500 columns listed in the PQRI database on that site -- you *had* to go and pick one that wasn't there!

All joking aside, I just did a quick Google search on hplc + c8 + "non-endcapped". This link cropped up on the first page:
http://www.chromspec.com/pdf/e/cs18.pdf

It lists the following as "L7" columns (C8). You would have to check them one at a time to see which are end-capped:
ZORBAX®:C8, SB-C8, Eclipse XDB-C8Eclipse Plus C8, Rx-C8
NUCLEODUR®: C8 ec, Gravity C8
NUCLEOSIL®: C8 ec, C8, C8 HD
Pinnacle™ DB C8, Pinnacle® II C8, Ultra C8, Viva C8
YMC-Pack Pro C8, YMC-Pack C8
Genesis® C8, APEX C8, Hx-Sil C8

Dear Tom,

thanks a lot for the list; I will check

Florian

things being very close then you should also check for a C-18 with no endcapped
now since the no end-capped is of interest, does it means that the OH interaction is going good to your separation?
have you checked against a end capped c-8? do you get worst resolution?
do your compound have some polar groups in them?

if the logic is sound to you, maybe you should look for a C-18 column with a polar group.
if the compounds come off too quick you could consider HILIC also. as well.
be carefull that the non-capped column you go for is not too "old". they generally are of older columns generations.
the worst is to have a method with a column that 2 years after is discontinued because it is old.

Dear unmgvar,

thanks a lot for your input.

I tried to separate the following two substances:



With Hypersil MOS-1 (non-emdcapped), I got a separation (R about 1.5)
With Hypersil MOS-2 (endcapped), the separation was worse
With Phenomenex Synergi Hydro-RP (hydrophil endcapped), I also got a separation with R about 1.5

Thus, my additional screenings will include hydrophil endcapped phases as well as non-endcapped C8-phases.

The mobile phase: AcN/H2O 22:78; Flow = 1.0 mL/Min; tR about 14 Min.

Regards

Florian

This is going to be a difficult separation on a standard alkyl phase, endcapped or not. As unmgvar pointed out you are using the -OH to help create the separation. My experience is the more "Type A" the silica, the better in these instances. I'd look at a phenyl phase. Maybe check manufacturers websites for something similar.

Quick question - have you evaluated other mobile phases (like Methanol:Water) with your column for these compounds?

Dear bisnettrj2,

up to now, I tried only acetonitrile

Regards

Florian

Haven't we had estriol as a topic recently?
The molecules differ in polarity (very little) and in steric shape (also very little).
Things like polarity of the column and hydrophobicity of the column a rather broad categories. In my experience with steroid impurities C8 and C18 differ a lot. Much more than Uwe Neue predicted from his measurements with a broad range of analytes.
I once had to separate a estradiolester from its impurities, especially 9,11-dehydro (similar to your second compound) and 6-dehydro (with a double bond in the lower right corner). I screened a number of columns with methanol and ACN gradients. There were 6 substances. In most cases only 5 peaks were visible, or 5+a shoulder. From the columns with 6 peaks in at least 1 or two runs I made Drylab models to find isocratic conditions.

Basic line- Try to find condition that fits. Column comparisons like on the USP site are helpful, however in this case not sensitive enough. Phenyl phase should theoretically add pi-selectivity, practically thats combined with retention loss. Temperature changes really can help. So far i don't see any rules that apply here - its trial and error.

alex

If you can, try using methanol instead of acetonitrile, and see if you obtain better or worse resolution. Different solvents produce different selectivities on similar stationary phases, and evaluating different solvents is part of good method development (as well as evaluating the effects of temperature).

It looks like a HILIC column could be an alternative. Some HILIC columns, were silica is modified with Carbamoyl group are not endcapped.

The compounds are phenols, as drawn, pH, therefore, should strongly influence their behavior. I suspect that the endcapped, nonendcapped diff. is due to some differences in charge exclusion. Since the double bond is in conjugation with the benzene ring there should be a difference in dissociation properties.

Dear all,

thank you very much for your inputs.

Just to let you know: the European Pharmacopoeia uses for the separation a NP-method using a Diol-Phase with Heptane/2-Propanol as the mobile phase.

Is this an indication that also for a RP-method a polar stationary phase should be used?

Regards

Florian