Orthoganol HILIC Method

Discussions about HPLC, CE, TLC, SFC, and other "liquid phase" separation techniques.

16 posts Page 1 of 2

I have a validated method for the determination of 22 known impurities for two active's. I have them all separated with the exception of 1 impurity that co-elutes with 1 active and two other impurities that co-elute with each other. The method is a gradient method using a Phenomenex Biphenyl column, 150mm X 2.1mm, 1.7um.

I have been able to develop a secondary method that is able to separate these peaks, (however other co-elutions exist). Nevertheless worst case I can properly identify/quantitate these peaks using a secondary method if needed.

My real question, is that for my own interest I wanted to develop an Orthogonal method to the Biphenyl column method. I've tried the following:

Primary Method:
Column: Biphenyl 150mm X 2.1mm, 1.7um
Mobile phase A: 0.1% TFA
Mobile Phase B: Methanol
Temperature: 35C

Alternate Method 1:
Column: HSS T3 C18 150mm X 2.1mm, 1.7um
Mobile Phase A: 0.1% TFA
Mobile Phase B: 75:25 Acetonitrile:Water
Correlation (Comparison of K' of Primary method to K' Alternate method): 0.80

Alternate Method 2:
Column: HSS T3 C18 150mm X 2.1mm, 1.7um
Mobile Phase A: 10mm Sodium Acetate buffer pH 5 containing 0.25% TFA/L
Mobile Phase B: 25:75 10mm Sodium Acetate buffer pH 5:Acetonitrile containing 0.25% TFA/L
Correlation (Comparison of K' of Primary method to K' Alternate method): 0.80

I can't seem to find a truly orthogonal method (Correlation coefficient as low as possible)

I wanted to try a HILIC method. I've never successfully developed a HILIC method.

I've tried using a Luna Silica (2) column 250mm X 4.6mm column with a mobile phase of 100% Acetonitrile, and 97:3% ACN:Water.

I've tried using a Luna Amino column 150mm X 4.6mm same mobile phase as above - Was able to get some retention, but nothing significant.

I've tried using a Kinetex F5 column 100mm X 2.1mm, 2.5um mobile phase 95% ACN: 5% 5mm Ammonium Acetate buffer pH 3 - Peaks elute in solvent front.

I've tried using a Phenomenex cyano 150mm X 4.5mm, 4um mobile phase 95% ACN: 5% 5mm Ammonium Acetate buffer pH 3 - Peaks elute in solvent front.

Some of the peaks are in fact extremely polar which is why in my original methods above my mobile phase A is 100% Aqueous.

Why can't I get significant retention on my HILIC columns?

Hi Geof235,

Please, what is the solvent you are using as the diluent for the samples in the HILIC work as compared to the reversed-phase methods? Remember that the ACN content must be approximately the same as that of the starting conditions of the HILIC method.

This may also help:

https://www.sigmaaldrich.com/content/da ... 132013.pdf
Hello again,

While I am at it, try this website, it may help you select other alternative phases that in principle should afford orthogonal separations to Phenomenex' Kinetix Biphenyl:

Hey Matt

Thanks I’ll check Those out

So my sample using the primary method is in 92:8 buffer:methanol

Before I inject the sample containing 20+ peaks I took my stock standard (containing only two peaks) and diluted down in 100% acetonitrile as you suggested just to try to retain/ separate the two peaks before I try something complex.

I don’t know if I’m doing something wrong or if these molecules just don’t retain on HILIC

I’ve seen some other chromatograms where the max run times are only 10 minutes? Is it just not possible to get anything to retain on hilic for more than this and I just won’t be able to separate such a complex mixture using this technique?
Just some further information,

The Luna amino column appeared to give some retention (about 6 minutes for both compounds) while all other columns gave retention of about 2 minutes seem to be in solvent front

Matt I did already find and read through he first link.

My other instinct is to try a weaker solvent. They say acetone is weaker but has a high in cutoff (330nm) which wouldn’t work for me.

Is there any other common solvents as weak or weaker than acetone that still share some aqueous miscibility (so I don’t have to convert my system to full normal phase) and so I can dilute my sample down to run on the orthogonal method. That also has a low uv cutoff as this will be an impurity method

Thanks again
Hi Geof235,

If you include the flow rate values for these different separations you're trying, it will then be possible for comment regarding retention. Off-hand, six minutes for a 4.6mm x 250mm column at 1 mL/min would likely not be enough retention (something more like 8.5 minutes would afford a retention factor of 2).

I think that it is certainly possible to obtain retention of analytes on a HILIC column for longer than ten minutes, at least based on the literature I have looked at. Of course, these papers likely do not include separations of analytes similar to yours in this situation.

A weaker-eluting HILIC solvent than acetone? Of the commonly-used HPLC solvents, the only thing that comes to mind is tetrahydrofuran, which is not an ideal choice.

If you decide to choose reversed-phase (if HILIC cannot be made to work), one suggestion for a stationary phase that would be orthogonal to Phenomenex Kinetix Biphenyl is Agilent's Zorbax Bonus RP, according to the hplccolumns database.

EDIT: Noticed that the amino column was 4.6mm x 150mm, if a 1 mL/min flow rate was used, then the retention factor(s) are > 2, though without a chromatogram it is hard for me to speak about resolution.
Hey Matt,

Thanks again all flow rates for the HILIC columns were 1ml/min..

So to give a little more background as well my mixture has two apis in it. One is a secondary amine with multiple rings and is a salt which does have some pretty decent retention on a C18 and phenyl column using mid to low organic compositions (90% - 80% aqueous). The impurities for this compound are structurally very similar and behave similarly. Likewise the other api is a very polar molecule with little to n retention on c18 or biphenyl elutes under 10 minutes using 95%-90% aqueous. One of the impurities for this compound has little to no retention even at 100% aqueous on both c18 and biphenyl (I can get it to separate from solvent front but barely). I have tried ion pairing but that didn’t help much either. I figured that the highly polar molecule would be ideal for HILIC.

One thing I did notice was that one of these papers mentioned better selectivity at higher phs (~9). Could this be because my molecules are weak bases and rendering them unionized would play into the dipole moments over the ionic separation?

The more I think about this it makes sense...
Another point... I found a method online that appears to be a HILIC method for a secondary amine structurally very similar to my polar secondary amine salt I previously mentioned using the following HPLC conditions

Luna silica (2) 250mm x 4.6mm, 5um
Solution 1: acetonitrile:diethylamine:water (950:4:1)

Mobile phase: solution 1:methanol (55:45)
Hello again, Geof235,

You said,

I found a method online that appears to be a HILIC method for a secondary amine structurally very similar to my polar secondary amine salt I previously mentioned using the following HPLC conditions

Luna silica (2) 250mm x 4.6mm, 5um
Solution 1: acetonitrile:diethylamine:water (950:4:1)

Mobile phase: solution 1:methanol (55:45)

Please, can you send a link to the method you speak about from Phenomenex, I would like to review it even though it does not seem to be either the problematic API or impurity. I want to be certain of the mobile phase conditions.

Generally, I have seen HILIC methods run at pH values between 3 and 6, at times with the introduction of a salt as well. On the idea of alkaline pH for a HILIC separation, I am not so certain.

There is a way to evaluate the logP values of the analytes you are trying to separate, too to see how well HILIC may work:

Hey Matt,

The method is the usp monograph for hydrocodone bitartrate api

The assay method specifically. I’ve run this method and it produces a peak at about 5-6minutes. My api is not hydro but it is very close structurally to that. The peak for hydro isn’t the greatest but that may be because the column is overloaded I’m also wondering if reducing the methanol comp would increase retention.

Plus I also found this from waters regarding HILIC method development.

http://www.waters.com/webassets/cms/eve ... 4FINAL.pdf
Good Morning,

Understood, please note that I do not have access to the USP. I was able to find an old copy of the USP Hydrocodone Assay method, the conditions are those you note above. The retention of the main peak is reasonable, agreed. It is certain that the injection solvent is more strongly-eluting in strength than the mobile phase based upon what I read--this and the 20 uL injection volume explain the poor peak shape, not column overload, per se.

Do you mean reducing the methanol content of the eluent? If so, yes, as methanol is a stronger eluent in HILIC than acetonitrile.

I will take a look at Waters' presentation. Interesting that they show alkaline conditions for HILIC, I would caution though to choose the stationary phase carefully for this type of application.
First of all, at high pH retention in HILIC declines for most classes of compounds on most columns (maybe Waters' BEH materials are an exception...). Since your retention is marginal at best, I'd avoid high pH.

Here's some guidelines for you that should serve to increase the retention of your compounds:
1) The starting mobile phase must contain at least 2-3% water. This provides the water needed to form the immobilized aqueous layer on the surface that is required by the partitioning mechanism.
2) Most reasonably polar small molecules will be adequately retained by a HILIC column with high capacity if you start with 90% acetonitrile (ACN) in the mobile phase.
3) The sample solvent should contain the same organic solvent as the starting mobile phase, and within 10% of the concentration. If your starting mobile phase contains 90% ACN, then your sample solvent should contain at least 80%, and preferably 85%.
4) It takes longer to equilibrate a column in HILIC than in other modes. Try equilibrating it for 40-45 minutes before starting a run.
5) Salt: Any silica-based column will exhibit at least some electrostatic properties. You need about 20 mM salt in the mobile phase to shield most of them. This is nicely illustrated in the slide set that you reach via the link in the preceding post to the Waters presentation. The electrostatic effects can be either attractive or repulsive, depending on the charge of the analyte and of the column (decidedly (+) for amino columns, and usually (-) for the rest, especially uncoated silica and the ZIC- materials).
6) Which salt? It turns out that the hydration of the counterion plays a critical role in determining the retention of a charged analyte in HILIC (ref.: A.J. Alpert, J. Chromatogr. A, 1538 (2018) 45-53). If your analyte is an amine with an ionizable nitrogen, then pairing it with a well-hydrated phosphate anion will increase retention quite a lot compared with pairing it with an acetate anion. While you can't dissolve much KH2PO4 in 90% ACN, triethylammonium phosphate (TEAP) is quite soluble in 90% ACN. Prepare this by mixing a known amount of H3PO4 with water and then adding triethylamine until you attain the pH that you want (3.0 would be good to try initially). Use this as a stock solution when preparing your mobile phase. If the starting mobile phase contains TEAP, then your sample solvent should contain it as well. If the sample solvent has a different salt than the one that's in the mobile phase, you can get really terrible peak shapes as some molecules exchange counterions during the run. The resulting ion pairs differ in polarity and migrate down the column at different speeds.

Hope this serves as a roadmap to get you started.

Andy Alpert
PolyLC Inc.
(410) 992-5400
I would like to separate fucose and rhamnose by HILIC, which seems to be problematic. I would prefer to not use the additives chloroform or phosphoric acid in the mobile phase, which are sometimes used with Polyamine II columns. I would like to use a CAD for detection, which does not tolerate phosphoric acid (and I think that chloroform is not allowed in the lab of health reasons). I prefer a method without derivatization.
I have thought about amino-propyl columns, which have the advantage of not giving double peaks of anomers of reducing sugars, even without addition of amines in the mobile phase at 25 C. However, amino columns are known to form Schiff bases with reducing sugars and sometimes give low recovery and problems with quantification and decreasing retention times.

I am grateful for suggestions from Andy A or other experts.

Best regards,

Goran K
AF Consulting
Goran's posting is somewhat unrelated to the subject of this thread, but I suspect that nobody who follows this Forum cares much about that...

This concern about Schiff base formation is, for the most part, a relic of the days pre-1990 when people had no idea what the mechanism was by which amino-silica columns retained sugars in mobile phases that were predominantly composed of acetonitrile. There was speculation that it involved transient formation of Schiff bases. That explanation ignored the fact that the method worked just as well with nonreducing sugars. Schiff base formation has been demonstrated in cases where marginal amounts of reducing sugars were left in amino columns for a prolonged period, circumstances that do not describe how people actually analyze sugars.

The other concern here is the steady loss in capacity by amino-silica columns. That's because the amino- groups create a local microenvironment of high pH. This causes the loss of the silane coating through hydrolysis, hence the decrease in capacity.

If you don't want to use a neutral HILIC column with 0.1% triethylamine in the mobile phase (which accelerates the rate of mutarotation, preventing separation of anomers), then consider getting a column of Shodex's NGP-50 material. This is a polymeric material with a tertiary amine as the functional ligand. The use of a tertiary amine precludes formation of Schiff bases, while the use of a polymeric material results in stability at elevated pH. The published applications with small sugars look good.
PolyLC Inc.

(410) 992-5400

edit: Shodex now calls this column VG-50.
PolyLC Inc.

(410) 992-5400

16 posts Page 1 of 2

Who is online

In total there are 18 users online :: 0 registered, 0 hidden and 18 guests (based on users active over the past 5 minutes)
Most users ever online was 599 on Tue Sep 18, 2018 9:27 am

Users browsing this forum: No registered users and 18 guests

Latest Blog Posts from Separation Science

Separation Science offers free learning from the experts covering methods, applications, webinars, eSeminars, videos, tutorials for users of liquid chromatography, gas chromatography, mass spectrometry, sample preparation and related analytical techniques.

Subscribe to our eNewsletter with daily, weekly or monthly updates: Food, Environmental, (Bio)Pharmaceutical, Bioclinical, Liquid Chromatography, Gas Chromatography and Mass Spectrometry.

Liquid Chromatography

Gas Chromatography

Mass Spectrometry