Polar-embedded selectivity

[JA]

Can anybody overview the statement that polar-embedded phases provide complimentary (or even "orthogonal") selectivity to a typical C18?

I thought polar embedded's were brought about for peak shape benefits only.

thanks

[Mark Tracy]

There are three benefits to the polar-embedded phases:

1. Improved peak shape for bases. This is somewhat akin to the base-deactivated phases, but the polar groups are non-ionizable and non-leachable (ideally). Properly done, you get improved performance for bases without sacrificing acids or neutrals. There are of course other approaches to well-shaped peaks for bases.

2. Aqueous compatibility. Most regular C18 phases are so hydrophobic that 100% aqueous mobile phases don't maintain wetting of the surface. This is called variously dewetting or phase collapse. <Insert dissertation here.> Polar-embedded phases have enough hydrophilicity to maintain wetting even in aqueous buffers. Again there exist other ways to accomplish this.

3. Alternate selectivity. This is hardly "orthogonal" to C18, more like a few degrees of tilt. Polar-embedded phases have multiple retention forces including hydrophobic, dipolar and hydrogen bonding. In the ideal C18 phase, the only interaction is hydrophobic, and anything else is a residual effect. Polar-embedded and other techniques attempt to engineer controlled levels of other interactions (they are now features not bugs). A truly orthogonal selectivity would be something like HILIC or normal-phase, where hydrophobicity is the bug and not the feature. The selectivity changes are highly dependent on the compounds in question and the bonded chemistry. Sometimes people use some kind of "polarity index" to compare columns, but these indices are only general guides. Just for instance, the PolarAdvantage column that Dionex makes acts more like a C18 column for most situations, with minor selectivity changes, except that it completely scrambles the elution order for EPA 8330 (alas, not in a useful way). I'm sure other vendors can cite their own examples, and will. If you root around on vendor websites, you can surely find example chromatograms.

I hope this helps and doesn't make your eyes glaze over...

[Uwe Neue]

Let me add to Mark's comments!

The improved peak shape for bases goes along with reduced retention for bases, since the influence of the surface silanols is largely eliminated. We believe that there is a shield of tighly bound water around the embedded polar group, which is why we called our packings "...Shield".

The embedded polar group also exhibits increased interaction via hydrogen bonding with some analytes. The most pronounced effect is the increased retention of phenols on a packing with an embedded polar group, but other functional groups that can hydrogen bond also have increased retention (e.g. sulfonamides, undissociated carboxylic acids etc.)

For a large set of analytes, we found that among all reversed-phase packings the ones with the embedded polar groups gave the largest selectivity differences to a standard C18.

[adam]

Uwe

Are you saying that these tightly bound water molecules are what keeps the analytes from getting down to the basic silica. I'm not sure I understand how that would work. Would you be kind enough to explain.

Thanks MAS

[Uwe Neue]

Think of it this way: if there is ice on the ground, it is difficult to get your shoes dirty

[Bill Tindall]

I tried one of these embedded columns for acids some time ago. The **#!!$ thing acted more like an anion exchange column. Aliphatic acids tailed badly and I was able to separate iodide, nitrate and nitrite on it with an ionic strength gradient (to demonstrate to the skeptical vendor that it was acting like an anion exchange column). Is this experience unique to this particular brand of polar embedded column? I never tried another brand.

I never do basic compounds. I see more or less evidence for anion exchange on some "modern" columns that I did not see on ODS II, for example. I suspect that with the emphasis on analysis of basic compounds that traces of protonatable sites have crept into column technology, probably to the delight of the pharma people, but its not nice for strong organic acids.

[Chris Pohl]

Bill,

I'm not so sure your observation is general. There are definitely some phases which have this problem in spades but tests with acidic solutes show that a number of commercial products (including, of course, the ones that Dionex produces) are free from these effects (although I can't say we necessarily tested them with iodide, for example). How widely did you survey these phases in drawing this conclusion?

[Uwe Neue]

Bill, I agree with you that there are a lot of phases out there which are as full of amino groups as an old bonded phase was full of silanols. But this is not the case for other phases, so we need to be careful not to generalize.

Many, if not most, of the embedded polar phases are amide phases that are created in a 2-step process. You bind an amine to the surface first, than you react it in a second step with an activated long-chain carboxylic acid, such as a C12-COCl. If you do that you have invariably some amine left on the surface that creates these effects that you have observed.

From my standpoint, this is the amateur's way of making such a phase. Companies with a real synthetic capability synthesize the desired ligand first, and then bond the complete ligand to the surface. In addition, in the case of the ...Shield RP phases by Waters, we use a carbamate linkage, with the oxygen of the carbamate oriented towards the silica. This ensures that even if there were a breakdown of the embedded polar group that you will not be left with a silly strongly interacting group on the surface. I have presented this case at more than one HPLC conference and in papers, and a few of my colleagues have improved their EPG product.

I have not tested the phases that Chris is talking about, but considering that they were introduced late in the game and have a complex design, I do not think that they made the mistake of the left-over amines that early phases and primitive copies of these early phases were suffering from.

As with standard bonded phases, there is high quality and low quality, and a knowledgeable user will find out quickly.

[HW Mueller]

Now, after all this work someone comes along and willfully adds an entirely different type of interaction capability (ionic), and someone calls this the greatest breakthrough (broke through what?). I don´t want to be misunderstood, as mentioned before, I have used some niche possibilities for multies, with if, if, if...... Otherwise: a pain in the neck.

[Bill Tindall]

Chris, Uwe others,

I did not know how general my observation was, which was the point of the question. I found your responses enlightening and they would encourage me to try other varieites of this column type. Most column manufacturers refuse to reveal the details of column chemistry like you have provided so the customer has no way to suspect things like incomplete amidation. (I had the main technical person from a major column manufacturer tell me that amines were not used in any way in the manufacturer of their products.) I later learned from a more knowledgable/honest? source that amines are used in both silica preparation as well as bonding and that these amines can have harmful basic impuirities.) The source of the column I tired, some time ago, "guaranteed" there were no basic sites left in the stationary phase after synthesis, which was shown to be incorrect. They probably didn't know. Which brought me to the second point, which I probably didn't make clearly.

The main market for columns seems to be with people doing basic compounds. It also seems that the details of column manufacturer can result in traces (ppm or less?) of functionality not revealed by the gross column description- "polar embedded, C18", etc. If the vocal customer feedback is from people worrying about doing basic compounds, things that affect the separation of acidic species can (do) creep into column manufacturing processes that aren't noticed by all but the rare customer. I don't think that column manufacturers generally have much experience with how their columns perform with difficult acidic compound separations(trace component or low pKa). Having been mislead on several occasions, I have become cautious about claims of column performance for acidic species and suggest that others be cautious too, if there is any one else in the world that does trace, strong acid separations.

If manufacturers want a test species to probe undesirable acidic interactions, I would suggest tricarboxy benzoic acid (trimellitic acid) near its detection limit. If there is too much hydrophobic "ballast" on the test species subtle effects may not show up. So choosing a test species by pKa alone is insufficient.