Selectivty of Neutral Compounds in Mixed Mode Separations

[Rob Burgess]

Currently we're trying to re-develop a simple isocratic method for three components: a neutral sugar ester, a basic pharmaceutical and a neutral steroid.

Previously I have looked at mixed mode chromatography before but did not achieve any selectivity whatsoever between the sugar ester and the steroid when used with either a 50,55 or 60% aq. AcN mobile. With reversed phase chromatography more than adequate selectivity can be achieved for the sugar ester and the steroid.

Is this generally the case then that selectivity of neutrals is diminished on mixed mode columns compared to RP? Is there anyway to enhance the selectivity? Ideally we're like to keep the method isocratic for ease of transfer purposes. Do most mixed mode columns commercially available offer similar selectivity for neutrals or do some offer better selectvity than others?

PS. The basic pharmaceutical is retained very well by changing the level of TFA in the mobile phase and is well resolved at a greater retention compared to both neutral solutes.

[SIELC_Tech]

Dear Rob,

Please contact Hichrom Ltd. in UK about free screening services we are offering using our Primesep mixed mode columns. We have 8 different mixed mode phase which might be suitable for your application. All of the column have C13 fragment with different ion-pairing reagent attached to the surface)

check our website for the applications developed for mixed mode columns (you can sort it by date, by compound and by column type). There are few applications for amines, amino sugars and neutral compounds, one of the brochures have application for steroids.

Contact us if you need more information.

[Rob Burgess]

As is the case for most pharmaceutical companies, due to the sensitive nature of our work / compounds we are not able to supply any of our compounds for screening.

We have tried the PrimeSep 100 column, and it was this column where we did not achieve any resolution between our neutral solutes. Is there any other mixed mode columns with a longer alkyl chain that may offer increased selectivity then?

[HW Mueller]

We discussed that before, my conclusion was that you can not have optimum performance with mixed mode. Of course, you may do allright with sub-optimum performance.

[SIELC_Tech]

Rob,

You problem is not in poor selectivity of the mixed mode column-it just happened that your particular pair of compounds has very similar retention on Primesep 100 column with selected mobile phase (TFA as modifier). In case of common RP chromatography you would have very few choices in this situation (change mobile phase, temperature, and organic modifier) which usually ends up in changing of the column. In mixed mode chromatography you have at least one extra option to change contra-ion that bounds to the ionic function on the column. On the Primesep 100 column you can screen this range of buffers: phosphoric acid, sodium phosphate, potassium phosphate, ammonium phosphate and triethylamine phosphate (10-50 mmol). The selectivity of the column towards critical pair will change drastically since the polar properties of the stationary phase change when the column goes from H+ to K+ or Na+ etc. forms. And of cause you can choose different stationary phases; although hydrophobic properties are similar on all Primesep columns the polar interactions are very different. We have 6 different RP phases in cation-exchange mode alone. Column efficiency typically is over 80K plates/m which is similar to any regular RP column. So the conclusion of HW Mueller “that you can not have optimum performance with mixed modeâ€

[SIELC_Tech]

Rob,

Check this link. It shows the effect of different buffers on the retention of xanthines, the same principle is applied to your application except you are looking at different cations in the mobile phase.

http://allsep.com/makeChr.php?chr=Chr_002

[HW Mueller]

SIELC_Tech, there is no inclination, here, to repeat the argumentation again, but surely I would like to point out that you and your colleague(s) have stated that you do not claim that your mixed modes do a better job than any of the normal RP columns on seps of neutral molecules. If I see this correctly the problem here is with the two neutral compounds. If I had such a problem I would start by looking for a column with the largest surface area and the least "non-hydrophobic" interactions. What specific surface areas do your columns have? How much is octadecyl (etc) of this area?

[Yury Zelechonok]

To HW Mueller
With the same chance on success you can go to more polar interaction or to different type of interaction like Pi-Pi. To your knowledge, any RP column is a unique combination of polar and non polar interactions. It does not mean that every column with two co-eluted compounds is underperformed in general. If two compounds are co-eluted you need to find a column with different set of those interactions, or alter the degree of one particular interaction by the MP. I can also argue that to resolve two neutral compounds with way different nature, but with the same retention on “goodâ€

[HW Mueller]

Yury, I already answered your question in a post some months ago, but you don´t answer mine, last one: how much area for ionic, how much for reverse phase of how much total. ...

[Yury Zelechonok]

Your question just indicates that you never even try to learn about Primesep phases. These mixed mode columns made with a single ligand which combines both an ionic functional group and a long alkyl chain - something like this Si-(CH2)n-A-(CH2)m-CH3 Where A is a weak cation-exchange or anion exchange group. The fact that the ionic group is placed within hydrophobic environment differentiates this phases from all other mixed-mode and RP columns. The carbon load on a column is from 11 to 13% and depends on the type of the Primesep stationary phase and pore size. Surface for 100A material is about 300 sqm/g fully cowered with one type of ligand. 300A material is available with surface area about 100 sqm/g. This link gives you more information if you are interested:
http://allsep.com/Technology_NovelStationaryPhases.php

[Kazimierz]

Dear SIELC_Tech,

Which Primesep column are resistant to heptafluorobutiric acid (HFBA)?
With Phenomenex Synergy Polar-RP HFBA 0.2% causes very efficient separation of ganciclovir.

[SIELC_Tech]

I believe that HFBA is used as ion -pairing reagent in this case. You do not need IP with Primesep columns it has ion-pairing reagent attached to the surface of the silica. You can use our coliumns with ACN-water and any of the following TFA, sulfuric acid, formic acid, ammonium formate,ammonium acetate or any other salt buffer. Safe pH range is 1.5-7.

Your compound is guanine based so here is the link to guanine retention on Primesep column:

http://allsep.com/makeCmp.php?cmp=Cmp_050

[HW Mueller]

SIELC_Tech, Yury

There was essentially nothing new there compared to last time I looked. I didn´t really want to know what you have on your website, rather, what is in your column.
Now, when I want an optimum RP column I am inclined to use what I bought: 100A pores with 450m^2/g surface (actually, a column whith even less polar interference would be welcome).
When more polarity seems adviseable the following is used: 80A pores 475m^2/g (probably less hydrogen bonding ability would be advantageous here).
When still more polarity, but not ions, are indicated this will do fairly well: 100A pores, 320m^/g (compromise near the limit).
When I want optimum ion exchange, the dangling of C-18, or whatever, in front is not wanted (for instance, the demo analysis of porphyrins with IEX is somewhat embarassing since I have to admit that unionized bilirubin, etc., can interfere via van der Waals, etc., even though we use a classical IEX column).
Now it´s probable that manufacturers spend millions to improve on the columns available in the early 1970´s, and I am still surprised at the lowered tailing etc….. Now, maybe mainly because of the high silica purification possibilities, one can go backwards and get something useful. I just believe that this "back to the not so good old days" with multi stuff is not a panacea. And for getting a little personal, in return: you do behave, usually, as if everything can be done with your multies, if one objects you seem to admit to limitations (your variations also point to the impossibility of packing all functions into one column).

[SIELC_Tech]

Hans,

What do you mean by "your variations also point to the impossibility of packing all functions into one column" What variations?

Regarding the surface area: what ever is available as a bare silica in 5 um, 100A has approx same 300 m2/g. We tried several different silcas and got the same results due to universal nature of our chemistry.

"as if everything can be done with your multies" -not everything, we never said that, but we made a lot of applications not achievable with the current sate of C18 (regular RP with out ion-pairing reagent). If you have a challenging application we can consider to run it and compare to whatever exists

Lets no blow this discussion out of proportions.

[MG]

Rob Burgess, did you try using methanol in place of ACN on the same column? If the mechanism is reverse phase, you would likely see different selectivity with methanol.