preparative loadability of an ion pair

[JA]

I have a secondary amine functionalised polyaromatic compound to assess for preparative work. For release testing analytically, it had been run isocratically at high organic composition with 0.1%TFA on a C18 with k~3. Something stuck in my head about the increased loadability noted for neutral vs. ionised forms of an analyte and I was wondering might increased loading of this compound be expected in TFA over a formic acid modified MP.
I'm not sure if what appears to be fairly decent retention at low pH is due to an ion pair with TFA, but if so could this act similarly as a formally neutralised analyte in high pH buffer and give me increased loadability?

For the high pH work, which would undoubtedly involve Xterra prep phase, I'm not sure which of the following components would be desired if the intended isolation will include rotary evaporation of the organic followed by freeze-drying the product out of aqueous soln. Let's say, pH 10 with ammonium hydroxide (maybe just a fixed composition of eg. 0.1% ammonia in water?), ammonium bicarbonate, acetate or formate.

[Uwe Neue]

You will definitively get a higher preparative loadability at high pH, say around pH 10. I have used ammonia to adjust the pH, because it is easily volatile, and one can use it at rather high concentrations without going to far into the alkaline. I have used 100 mM ammonia without difficulty.

You can use ammoniumbicarbonate too, but the concentration is lower. This is why I preferred ammonia. However, if for whatever reason you want to use ammonium bicarbonate, you need to go above 60 degrees to evaporate. This may not be suitable for some compounds.

The loadability under acidic conditions is always less in my experience, but it also depends on both the stationary phase and the buffer. Some columns give easy overload (=tailing) in TFA, and are good in formic acid, some work the other way round.

[JA]

Uwe, have you simply added conc ammonia to water to get pH 10? Also, when we say 100 mM ammonia, is that actually 100 mM NH4OH, equivalent to 4 ml "880 Ammonia" per litre?

Unfortunately, I couldn't get any selectivity between the sample components on Xterra at pH 10. The critical difference is an alkene or alkyne function bridging two benzene rings, the sample comprising both alkene stereoisomers and the alkyne.

[Uwe Neue]

I am not familiar with the designation of "880 Ammonia", so I can't help you to calculate how much of the stuff you need.

Sometimes you can get changes in separation selectivity for such a pair by switching from methanol to acetonitrile. On the other hand, to go down this path means work, and you may not wnat to do that.

[JA]

what do you mean I may not want to do that. Unlike the extremely lucky German Analytical Chemists who have a hundred robots performing day-to-day tasks in the lab, I had already gone in today to try these things out myself. Luna C18(2) and Xterra C18 don't look hopeful for the separation, while Kromasil C18 is showing good selectivity for the double/triple bond in both formic and trifluroacetic acids. Similar k in both those modifiers (maybe no TFA ion pair?) however, TFA seems to offer additional selectivity between the alkene isomers. I just hope I get plesantly surprised when I try to load up on the Kromasil column..

"880 Ammonia" is, strangely, 0.88 SG concentrated ammonia solution (~35 wt%). What do you use for ammonia and how do you get 100 mM?

[Uwe Neue]

I know nothing about buffers. I just buy them from Buffers-R-Us

[Victor]

JA,

First look very carefully on your bottle. Let us assume it says 35% ammonia by weight (not 35% ammonium hydroxide by weight-don't think so but I don't know which country you are in).

Look carefully for the specific gravity on the bottle-if it says 0.88 this means that 1 litre of the ammonia solution weighs 880g.

If 35% of this is ammonia by weight, this means that 1 litre contains 308g.

Taking NH3 = 17, this means the stuff is 18.1M.

Of course 1 litre will contain 18.1 moles of ammonia and also 18.1 moles of ammonium hydroxide.....

These figures may vary if the bottle is an old one. You might get batch data from the manufacturer, or even a standardised solution, but I say this only for the purists. Precision is more important in HPLC.

Mine's a vodka-of course!

[JA]

Thanks Victor.

After some checking, they do list 0.88 SG ammonia as 18.1 M (18.1 N). Stupid thing is, from the one supplier, Fisher Chemicals, we have Ammonia solution, about 0.88 SG with a lot-to-lot specified SG (0.884 for the bottle currently in our lab) and a Guaranteed analysis of "about 35%" ..haha. I'm not sure if I should just assume its NH3 content doesn't change much and dilute down from 18.1 M. Strangely, I can't seem to find any decent purity aqueous standardised solutions around.

Out of interest, what ammonia solutions you got in your lab Uwe? Manufacturer, grade. What about the Sepsci guys?

cheers

[Uwe Neue]

We use Aldrich, and it says roughly the same on the bottle as what is written on your bottle.