2D HPLC: HILIC / HILIC hyphenation

[carlo.annaratone]

After attending the recent ISC 2014 symposium, I started updating my knowledge on 2D HPLC. One of the topics of debate is orthogonality: a high ortogonality is desirable, therefore the coupling RP / HILIC would be the best from a theoretical point of view. However solvent miscibility problems often arise, therefore RP / RP is way more common. To achieve some degree of orthogonality, different pHs or stationary phases are used in the two dimensions.

We all know from experience that RP phases cannot be that much different (http://pubs.acs.org/doi/full/10.1021/ac050923i). However different HILIC phases have extremely different selectivity (http://www.sciencedirect.com/science/ar ... 731201895X). So why not couple two HILIC columns for 2D HPLC of polar compounds? It seems a promising idea. Hower I could only find two papers : http://onlinelibrary.wiley.com/doi/10.1 ... 900086/pdf and http://www.sciencedirect.com/science/ar ... 541200214X

Are there more studies? Can anyone point me to papers? Of course any discussion on the topic is welcome

[Gerhard Kratz]

We all know from experience that RP phases cannot be that much different (http://pubs.acs.org/doi/full/10.1021/ac050923i).

If that would be right we would have 99,99% less problems in RP Chromatography!
Different base silica, different bonding compounds etc. and have a look on the hydrophobicity tables published in books, comparing RP columns.

I came late to ISC in Salzburg and missed that presentation, but I attended the short course at HPLC in New Orleans about 2D Chromatography presented by Dick Henry.
In my opinion 2, or more, different separation modes make 2D or multidimensional Chromatography.
To cut a peak from separation 1 on RP column A and inject it on RP column B with a different selectivity is an excellent idea, but not 2D. The same is with HILIC or IEX.

[DJ]

This depends on what is sought. An effective 2D-HPLC for proteins/peptides (as in the Anal Chem paper) may or may not be useful for small molecules (as in the other cited papers).

A pH 2, pH 10 2D RP-HPLC makes sense for peptides/proteins where the goal is to ID as many individuals as possible from a complex mix. There is no shortage of papers detailing such approaches (SCX-RP, RP-RP,- HILIC-RP, you name it) 2D HPLC in proteomics.

In the HILIC-HILIC 2D HPLC references you cited, polar solutes were sought. My guess is that HILIC-HILIC 2D HPLC would be an effective alternative for non-ionizable solutes that are too polar for RP-HPLC.

In general, I tend to think there are more ways to exploit different RP phases (cyano, PS-DVB, C18, pentafluoro, etc) with different mobile phase additives, pH, type of organic, for orthogonal 2D than HILIC-HILIC 2D. If the compounds of interest are separable on RP-HPLC, it's hard to imagine this (RP) not being one of the dimensions.

[M_Farooq]

That is an interesting idea. HILIC columns show interesting selectivity changes when the stationary phase bears charged functional groups. Unfortunately, there is a big challenge in designing HILIC phases for orthogonal selectivity. It is a not-so-trivial exercise to prepare a good quality stable HILIC phases. One may think of many novel chemistries, but you can recall that HILIC requires a stagnant water layer on the silica surface. The high affinity of water to HILIC phases cause the bonded siloxane to hydrolyse and bleed from the surface. As a result, you are left with littles choices for a stable HILIC phase - hence limited choices. Although bare silica is also marketed as a HILIC phase, it is not free from problems of retention time drifts.

Y. Guo did a study on the selectivity of 8-9 HILIC chemistries for nucleosides under identical conditions. It turned out the elution order was mostly the same on all such silica phases, the retention times were different though. For orthogonal HILIC, a different material, other than silica might be a good combination in future for 2D HILIC. I developed some carbon based HILIC phases which showed different selectivity from 36 other HILIC and reversed phases but carbon has its own issues like lower efficiency than silica.

[carlo.annaratone]

If that would be right we would have 99,99% less problems in RP Chromatography! Different base silica, different bonding compounds etc. and have a look on the hydrophobicity tables published in books, comparing RP columns.

From one point of view, you're right, Gerhard. I do know that not all C18 phases were not born equal! However I do not think that changing from one C18 to another will ever bring to a total reversal or total randomization of the elution order. From what I've read and seen, the most orthgonal phases are C18 and pentafluorophenyl. Any good citation of literature is welcome.

This depends on what is sought. An effective 2D-HPLC for proteins/peptides (as in the Anal Chem paper) may or may not be useful for small molecules (as in the other cited papers).

Also this is true. The paper shows orthogonality of columns with a peptide mix. This does not imply that with another kind of analytes (whether it can be antibiotics, metabolome or any kind of small molecule) the orthogonality will be the same.

n the HILIC-HILIC 2D HPLC references you cited, polar solutes were sought. My guess is that HILIC-HILIC 2D HPLC would be an effective alternative for non-ionizable solutes that are too polar for RP-HPLC.

Yes, that's the whole point of the post! For non polar analytes, you can do RP-RP, for medium polarity you can do RP-HILIC, but nobody has explored yet HILIC-HILIC. What I was asking is:

- References from literature, if someone can find others
- Discussion about how one could achieve high orthogonality in HILIC.

Regarding the last point it seems obvious that one won't go very far using two bare silica columns. However there are many phases beyond bare silica, and there seems to be more difference in selectivity between them that between any tow RP phase (see http://www.sciencedirect.com/science/ar ... 7311008491 and http://www.sciencedirect.com/science/ar ... 7312012988). Ideally one would couple two colums that are opposite in the graph, obtaining high orthogonality, while still having the advantege of solvent compatibility (no peak distortion at injection, no sample dilution)

[Andy Alpert]

Here are a couple of more references for you:
1) Y. Wang et al., J. Chromatogr. A, 1181 (2008) 51
2) Y. Wang et al., J. Sep. Sci. 31 (2008) 1564.

These papers explore the use of HILIC x HILIC for analysis of polar compounds in traditional Chinese medicines. In addition, Pavel Jandera has written reviews on comprehensive chromatography that touch on the subject.

[carlo.annaratone]

Thank you Andy!