Different dilution solvent, different peak area?

[justo]

During the dissolution analysis of Digoxin tables we found that if the final STD is diluted (final dilution) with mobile phase has an area of (for example) 90000, if we make the final dilution with water the area is 66000, the Rt and the shape of the peak suffer no change. The concentrated std is prepared in methanol (drug with very poor water solubility). We made many test but we can´t find and explanation for this difference. Has anyone suffered something like this?

[tom jupille]

I can think of a couple of possibilities, but without more data (and more experimental work ) these are no more than wild guesses.

1. Are you *sure* the peak shapes are identical if you overlay the two traces. Depending on your integration parameters, a subtle difference in peak shape can make a big difference.

2. Given low solubility in water, is there a possibility of loss due to precipitation (the final solution is perhaps a bit hazy?) or adsorption onto the glassware?

[justo]

Dear Tom thanks for your ideas but the shape and Rt are identical and I sonicated for 30 + minutes the aqueous solution of STD trying to dissolve more digoxin and the resulting area was identical to the previous one (no more digoxin was dissolved) this discouraged me.
Really it is a rare fact for me, with more than 25 years working with HPLC.
Any further help will be appreciated

[bhuvfe]

Justo,


Just few thoughts.

How many ppm of Digoxin correspond to an area of 66000 (or 90000)?


Isn't that possible that you have simply reached the solubility limit
of this product in water at the chosen temperature?

As Tom pointed out, an adsorption effect can easily explain what you
have and why you have a higher area when using the mobile phase
(containing ACN or MeOH I presume). And if the concentration of your
analyte is low enough will be difficult to see this by eye.

Do you sonicate the solution of digoxin in the mobile phase too?
Or only when you use water?


Regards,
bhuvfe

[HW Mueller]

One can imgine that sonication might give supersaturation, but that can be reversed extremely fast, so part could quickly precipitate or stick on the container as mentioned above. That is, you can not really lift physical laws with ultrasound.

[Bruce Hamilton]

I not a big fan of guessing, so I would perform some more experiments - if you haven't already performed these.....

My suggestion would be to add some of the other components of your mobile phase to previously-analysed water-diluted sample solutions - to an identical composition to the mobile phase - including any buffers etc.

I would warm and/or ultrasonicate those new solutions and analyse them.

If, after correction for the extra dilution, you recover the missing area counts, you need to ensure future dilution are with mobile phase.

If you still don't obtain the correct araes, then I would be concerned about the use of water dilution, which may have caused precipitation, sample degradation,
etc. etc.

I would also consider diluting and ultrasonicating and/or warming duplicates of some samples with twice the amount of water or mobile phase, to ascertain if the results are half the peak area.

Bruce Hamilton

[mohan_2008]

Considering the fact that the drug is poorly water soluble - I would not recommend it as a final dilution solvent - unless you know the threshold solubility of the drug in water.

Moreover, in my experience, I find that using a pretty strong mobile phase (say 80% organic) with a weak diluent (100% water) is not a good method - and prone to unprecedented difficulties.

The use of incompatible mobilephase can affect the analyte solubility as well as injection errors.

I would use the mobile phase as the first & final diluent if given a chance.

[HW Mueller]

mohan, what "unprecedented" difficulties do you see?
I have mentioned this before: One can inject mL amounts using a very weak solvent, without loss of resolution.

[mohan_2008]

I had method problems -

validated methods with water as the final diluent and the mobile phase very organic (80% organic modifier).

Technically, I agree with you - water is a weaker solvent and should pass along. But, we had problems with two such methods. Lot of investigations and filing LIR's but to no avail, as we cannot source out a specific reason.

[danko]

Could it be the methods that utilized 80% organic, weren’t RP-methods?
HILIC? If so, then the water is the strong solvent/eluent.

Best Regards

[Kimico]

I wonder if the difference in area counts you have experienced for your STD has to do with zone dispersion effects due to the solvent-solvent interactions from the diluent you used.

[HW Mueller]

Kimico, how would zone dispersion, whatever that is, cause a lowering of area?
Mohan, you don´t have to know how it happened, you can tell us what happened.

[Kimico]

Differences in chromatographic efficiency due to the injection solvent used (i.e shorter and broader peaks vs taller and narrower peaks).

[Uwe Neue]

Peak area is not affected by dispersion, at least not with a concentration-proportional detector, and for circumstances where there are no changes over the peak elution that would affect the spectrum.

[mohan_2008]

No, it is not Hilic.

It is a regular RP method, utilizing 80% Methanol and 20% Water.
This is a sponsor method - and at the time, we were just allowed to demonstrate a successful method transfer. Absolutely, no modifications.
The analyte was very hydrophilic and hence no solubility issues.

Several problems seemed to arise.

Inconsistent peak areas to start with. Standards fail system suit, if they pass in the long run, a bracketing standard would fail.

In contrast, we had a few recently developed methods that use about 30% organic only with 70% water. They work incredibly well with pure water diluent. Probably, the mobile phase being compatible with sample diluent in this situation.