Chromatography Forum

Hi there,
I am trying to reproduce a RP-HPLC method protocol a colleague gave me, using a Agilent 1200 system with binary pump. As organic phase, 60% Methanol + 0,25% Octanol is used. I have severe pressure problems and as far as I get it right, the PTFE frit in the purge valve seems to be clogged by the Octanol. Can anybody help please???
Thanks in advance,
Klaus

What happens if you just leave the octanol out?
Does the octanol dissolve easily in this MeOH/H2O mix?

When I leave the Octanol out, my separation isn't working anymore. so I really need that.
Octanol easily dissolves in Methanol and there is also no precipitation when mixing that with the aqueous phase.

If the octanol is freely soluble in the mobile phase, I would be surprised to see it plug the frit -- unless the octanol is somehow swelling the PTFE (???).

The easiest thing would be to replace the PTFE frit with stainless steel (if it's available), or leave it out altogether (make sure you have a stainless steel in-line filter downstream, though!).

I recall a method like this for determining log P by polycratic HPLC.

Do you use octanol-saturated water?

@tom jupille: Yes, the mobile phases are freely miscible. I also tried to run the system with bi-destilled water instead of HEPES buffer just to make sure that there are no salts to crush out. So octanol being - somehow - incompatible with PTFE is the only thing I can think about.
I have thought about leaving the frit out as well, but was afraid that I might ruin the system elsewhere. I should will try out a stainless steel downstream filter: that's a metal filter frit placed right after the pump, correct?
@ DJ: Yes, well, the aqueous phase is saturated with octanol by adding 10% of octanol to the HEPES buffer, stirring it over night, putting it into a separation funnel for four hours and then using the lower layer as mobile phase. The assay is used for measuring LogD by retention time / capacity factor of the sample compared to a set of standard substances.

Tom already talked about what I was getting at.
In your first post you talk about 0.25% octanol, in your last one there is 10% octanol, added in a way that you don´t really know how much is finally there. (I also wonder about the repeatability.)
Also what do you mean with clogged? Stop flow or just restricted flow?

Sorry for the confusion. I am using a binary pump so I am mixing those two buffer in the system:

aqueous eluent: to 1L 20mM HEPES pH7.4, 100mL octanol is added and stirred over night. The phases are then separated after standing in a separation funnel for four hours and only the lower HEPES phase saturated with octanol is used as aqueous phase

organic eluent: 2.5mL octanol is added to 1L methanol

The system is then run with 35% aqueous phase / 65% organic phase. So everything should actually be reproducible.

The flow gradually rises. Normal system pressure is about 110bar with column (Supelcosil LC-ABZ, 5cm * 0.46cm, 5µm) first, but then gradually rises within about 30min to about 360bar and a bit later to the system limit of 400bar.

I do not see any precipitation on the filter frit, it just smalls like octanol...

When I then purge the system with MeOH / water, the pressure only very slowly descreases, even with the purge vent open it is still at about 250bar. When I change the frit though, the system pressure goes down to 1bar with the purge valve open right away.

Have you tried to first saturate pure water with Octanol and use this water solution to make a buffer?BTW,this is the best way to do at least with phosphate buffer.
Have you filtered that saturated Hepes buffer?

I think I will pre-mix aqueous phase and organic phase and then filter that.

I should will try out a stainless steel downstream filter: that's a metal filter frit placed right after the pump, correct?

Correct!

I was wondering about the reproducibility of the octanol saturation.
Also, one can probably smell octanol on a number of items in your lab, not just on the frit.
What are you using to filter the mobile phase?
Finally, is this old hat that I have the inclination to think that chromatography is a bit daring for doing these distributions?

It's most likely to use octanol for solubility concern. if you change methanol to ethanol, you may be able to use less or no octanol. Just a thought.

[quote="immergruen" DJ: Yes, well, the aqueous phase is saturated with octanol by adding 10% of octanol to the HEPES buffer, stirring it over night, putting it into a separation funnel for four hours and then using the lower layer as mobile phase. The assay is used for measuring LogD by retention time / capacity factor of the sample compared to a set of standard substances.[/quote]

I have a little bit oxperience with logD by hplc. By your inclusion of HEPES, I assume you are interested in logp in the physiological pH range. Many (but not all) basic compounds will be ionized at pH 7, however, some (imidazole-like basic groups) are not. This means long long retention times if done isocratically.

An approach I used successfully (and have -never- added octanol to my mobile phase) is to perform several isocratic runs. Plot log k vs % organic. The y intercept from this regression is logKw, the theoretical retention factor under neat buffer conditions. The other regression coefficient is S (slope obtained from linear regression.

This means a lot of runs to determine a logKw (approx related to log P), and if your already messing with sticky octanol, might as well do the conventional shake flask, eh? :_

Another way to estimate logKw and S is from retention data of two gradient runs (change tg or sub MeCN for MeOH). From this, you can approximate obtain Kw and S numerically.

With these two properties on hand, it is theoretically possible to predict retention times for ANY gradient program.

DJ, you are saying that the retention time depends only on the dissociation constant?
(Even less sure about some of your other statements, but too lazy to check into it).