Chromatography Forum

Brian,

I have hands-on experience in designing and manufacturing of 19 commercial true mixed-mode phases (not blends of two or more silicas) I don't need any information from manufacturer on what is exactly on the surface of mixed-mode silica. I can make a lot of conclusions based on methods, mobile phases, peak shape, retention patterns, etc. Based on your applications I know that it is not pyridine or a weak base with pKa below 7. I am not sure that chemists who made Scherzo SM-C18 in Japan revealed the structures to people in US

When using the porous graphitic carbon, have you tried eluting with a steep gradient containing high % of acetonitrile and TFA or HFBA? Although if your compound did indeed contained aromatic rings maybe a phenyl column might be a good compromise...

Kostas> Good point! I have just tested a Zorbax SB-Phenyl with 2% ACN in 0.2% TFA. I get k' = 1, but the peak shape is not nice. At least some retention for the first time on a RP column.

I used 0.2% ACN in ACN as my eluent for the Hypercarb. That column is more like a mobile phase stripper. I look forward to test the weak anion exchanger suggested by Vlad.

Mattias,

If a Phenyl column works, perhaps considering a PFP phase wouldn't be a bad idea!?

Problem solved with the Obelisc N column, using a pH gradient from 5 to 7. I guess this pH increase switches off the anion exchange, releasing the sulfonic acid.



Looking at the retention compared to other sulfonic acids, this unknown molecule should contain 3-4 sulfonic acids. No wonder it was never released from ordinary anion exchange columns. So now I have a good method to hook up to the MS (mobile phase only water, 10-20 mM acetate and acetonitrile).

Nice (and thanks for sharing the outcome!)

What happens here is that Obelisc N has a strong acid and a strong base on the surface. (phase is zwitter-ionic). When you are increasing pH from 5 to 7 you are increasing ionization of the acid on the surface of our Obelisc N column, this acid will start neutralizing basic properties of the stationary phase, thus reducing anion-exchange interaction between basic group on the surface and acidic analytes.

I also want to mention that increasing pH reduces strength of the base to some extend, although strength of the analyte (acidic properties) is increasing too. In this case you see a complex interactions/equilibria which are impossible on other columns: in case of Sequat HILIC column you have both strong acid and strong base on the surface, but both groups are too close to each other to allow both cation-exchange and anion-exchange mechanism. In case of Scherzo, both groups are too far, because they are on different silica particles. In case of anion-exchange columns you need to overcome multiple anion-exchange interactions. This approach can be used in opposite cases, when compound with several basic groups is analyzed.

This chain has been quite confusing, just some examples:
Mattias, are you still assuming that this unknown is a sulfonic acid or do you know?
If one can´t use salts which elute anions from an "ordinary anion exchange column" wouldn´t one try a weak anion exchange column? Is that actually what you have done, in a complicated way, by using Obelisc N??
Vlad: Did you want to say that the column has WEAK acid and base substituents?
You go toward a neutral mobile phase but the higher dissociation (really due to increasing mp pH) of the acid substituent neutralizes the surface? This reduces anion exchange?? Are you not just decreasing the + sites by increasing mp pH?
How does increasing pH weaken the base (on the surface?) and increase the acidic properties of the analyte? Now, not seeing the role played by any negative substituents on the surface, I can´t understand this role of distance between + and -, except if you are unfortunate enough to have a situation where you may get ion exchange and ion exclusion at the same time.
I understood the reason for the zwitter on ZIC-HILIC to be for the purpose of having an alternate polar phase for HILIC. The interaction of ions with zwitters seems indeed very complicated.

I still don't know that it is a sulfonic acid, but I cannot think of any other acidic funcition in an organic molecule that will stay dissociated at pH close to zero.

I am using weak anion exchange SPE to isolate the impurity. In order to elute the molecule from this phase (Oasis-WAX) I need to go to 10-15% HCl in MeOH. I would never see it again if I injected it on a WAX column.

Regarding the theory behind the Obelisc separation, I must admit that I don't understand everything. I assumed that the positive charge of the phase was lost when pH was increased, and since the molecule has no RP-retention it would come flying out. Vlad wrote his more scientific explaination above.

For me it doesn't matter so much, since I now have a MS-friendly method.

Why would one use a weak anion exchange and then use acid to get the ion moving??
The weak anion exch. columns I have seen would spit out any (-) charged material at a pH of 5 or above, just like your Obelisc N.
Also, from the data you gave I wouldn´t dream to assume, as yet, that I have a sulfonic acid, let alone a multi . . . .

Good discussion.

Mattis,

Are you coming back to tell us what the interesting unknown is? Since you are run MS, which should help ID, right?

Hans I am glad that you can make conclusions even without knowing what is in formulation Mattias has and what are protential impurities/degradation products

My conclusions are based on our chemistry, experiments in the lab, and based on list of compounds in formulation. We are speculating what is happening and what are side products of main component of the mixture.

There is no problem eluting monosulfates from Obelisc N column at pH 3 and 4 (benzenesulfonic acid, pTSA) for other sulfates we had to increase pH.

The following experiments were conducted for Mattias (free of charge ). To show how retention time for strong acids is controlled by pH of the mobile phase.

Obelisc N (4.6x50 mm, 1 ml/min) runs showed:

1,5-naphalenedisulfonic acid retention at 40% ACN and 10 mmol buffer:
At pH 3 retention is 22 minutes
At pH 4 retention is 10 minutes
At pH 5 retention is 3 minutes

4,5-dihydroxynaphalene-2,7-disulfonic acid at 40% ACN and 30 mmol buffer:
At pH 3 retention is 7 minutes (longer with 10 mmol)
At pH 5 retention is 2 minutes

Congo Red (dye with 2 sulfates and two aniline fragments) at 40% ACN with 100 mmol buffer
At pH 3 no elution (over 30 minutes)
At pH 4 retention time is 28 minutes
At pH 5 retention time is 7 minutes
At pH 7 retention time is 2.5 minutes

Retention time is affected by number of sulfates, position of the sulfates, pKa of the sulfates, etc.

P.S. and yes Obelisc N has a "weak" acid on the surface with pKa of 3, along with the basic group

The data don´t confuse me, its the interpretation given in the contribution of Jun 10, 11:52am.
But, I am now even more confused as I thought I was the one who cautioned, rather than "make conclusions even without knowing . . . "

Why would one use a weak anion exchange and then use acid to get the ion moving??
The weak anion exch. columns I have seen would spit out any (-) charged material at a pH of 5 or above, just like your Obelisc N.
Also, from the data you gave I wouldn´t dream to assume, as yet, that I have a sulfonic acid, let alone a multi . . . .

The protocol for Oasis-WAX is to elute the retained compounds with diluted ammonia. That is not possible for me since the my unknown molecule is instantly degraded in that solution. It is only stable in very acidic solutions, that is the reason for eluting with acid instead of base.
Using acid, not even conc formic acid moves the compound. I need HCl. That suggests an extremely acidic molecule. The impurity is only formed when the product contains pyrosulfite (used as antioxidant). That also points towards sulphur being involved here. I hope the MS data will confirm/reject the sulphur theory (looking at the isotopic pattern)

Regarding the Obelisc column, I don't know of any other anion exchange LC column that could do this "switch-off-trick". It is not offered by Waters or Agilent at least. As Vlad wrote, he had been very helpful running several experiments in his lab. And this for an one-shot application (once I have the MS data, this separation will probably never be run again). Quite impressive service I must say.

Mattias, you are eluting at pH 5 - 7 on the Obelisk and it doesn´t kill your unknown? The same (or similar) conditions don´t elute it on the Oasis WAX? Any "weak" anion exchanger has a "switch off" capability, so what do you mean with this "switch-off-trick"? Does the Obelisk form more negative entities on its surface during the pH gradient, thus giving some exclusion (the acetate is not concentrated enough to obliterate that?)?