Dehydrocholic Acid Derivatization

[Sentenced]

Well, I´m new in this forum, and I want to know if you could help me a little...
I´m currently working on a method development, I must analyze dehydrocholic acid in tablets, but I have two main problems:

1 - I work mainly with HPLC/UV, and the dehydrocholic acid does not absorb in the working UV range.

2 - The matrix includes an artichoke extract, which gives me a lot of unwanted peaks on the chromatogram...

What I´ve been trying so far is the derivatization of the dehydrocholic acid with DNPH, in the following way:

1 - The sample is dissolved in MeCN:H2O (90:10) (about 0.04 mg of DHC per ml)
2 - The DNPH reagent is prepared dissolving 2 g of DNPH in 2 ml of H2SO4, completing the volume (100 ml) with MeCN:H2O (90:10)
3 - The derivatization procedure is: 4 ml of Sample + 1 ml of DNPH reagent, 120 min. at 60ºC, completing the volume with MeOH:H2O:AcN (83:2:15) (so the final proportion of solvents equals the mobile phase)

I am using a Kromasil C18, 25 cm, and I can integrate a stable peak, sensitive and linear to the concentration changes, but I can see a lot of other peaks that keep on growing with the time. I think that my derivatization procedure is failing somewhere, but I´m not used to work with DNPH... can anyone help me with this, or at least sugest any other way to analyze the DHC???
Thank you all...

[Mark Tracy]

DNPH reacts with any ketone or aldehyde, including sugars. Since you are not complaining about your main peak changing over time, I suspect that some matrix components (polysaccharides?) are hydrolyzing under the acid conditions to form reactive ketones and aldehydes.

As long as your recoveries and linearity don't change over time, you can ignore the other peaks. If it really bugs you anyway, you probably need to find some selective extraction for the dehydrocholic acid prior to derivatization.

[Sentenced]

Thanks for your answer

Maybe I didn´t explain it well...
This problem that I have, it happens even with the standard DHC, that is, without any matrix present.

I´ve derivatized 7 samples changing the time I heat them at 60ºC after the adding of the DNPH reagent (10, 30, 50, 60, 80, 100, 120 minutes).
I have two peaks, the one that I´ve found to be linear, possible the DHC+DNPH (peak 1) and an unidentified one. I´ve found that the peak 1 increases its area and becomes stable after 60 minutes at 60ºC, but the peak 2 goes from 7% to 120% of the peak 1 area. I think it´s important to note that the peak 1 area is, from 60 to 120 minutes, stable, and that this peak doesn´t appear in the reagent blank.

Any idea?

[HW Mueller]

What do you get in the derivatizing reaction if you leave the cholic acid out, everything else being kept constant including the amount of sample solvent?
In our hands this derivatization of ascorbic acid proved extremely messy. There were several peaks (not due to ascorbic) whose quantities were irreproducible and which were unstable. The derivative of ascorbic was an osazone which existed as two isomers that interconverted at room temp. in solution. It was decided that the old sugar chemists had enough fun with this reagent and that we didn´t need it.

[Sentenced]

You are right, I thing that I should look for another method to derivatize this particular analyte...
If I leave the cholic acid out, everything else being kept constant, as you said, I have a big reagent peak... and nothing else... but if I inject the sample, I have TWO peaks, the ones I´ve talked about before.
Well, I´ll try to look for another reagent, thanks anyway!

[Mark Tracy]

Dehydrocholic acid has three ketones, so you can (and probably do) get multiple derivatives. Yes, HW is right, each hydrazone comes in cis-trans isomers.

If you can strictly control the times and temperatures during the reaction and between reaction and injection, you can make this method work. If you can't be assured of strict control, then you should look for another reagent. (Too bad you are not looking at deoxycholic acid; there is an enzymatic method for that.)

[Bruce Hamilton]

If you search on dehydrocholic and HPLC on Pubmed and the WWW, you will find a couple of papers with alternative derivatisation reagents.
I take it the USP titration method for DHC tablets doesn't work?

My limited experience with DNPH was that trivial amounts of acid were required, and you may lose some fo the degradation by greatly reducing the H2SO4 content. Play with your standard, and then spike with some of the maxtrix ( which sounds delicious ) to ascetain if the method is worth pursuing.

Please keep having fun,

Bruce Hamilton

[Sentenced]

Thanks for your answers...
I`ve recently conducted a new assay, I`ve monitored the area of the two peaks along the time (20, 40, 60, 80, 100 and 120 minutes) at 80ºC, and I`ve found that the first peak reaches, at 20 minutes, a stable area, and the second peak keeps on growin until 60 minutes, were it stays stable too. I know what you`ve said about derivatives and isomers, and I though so too, but it`s strange that the first peak keeps its area, but the second one grows. I mean, if the second one is an isomer, a more or less substituted form or a degradation, the first peak shouldn`t keep its area value, right? (And I`m talking about a RSD% lower than 0,5%)...
The first peak is, as far as I know, linear, right, but I should do something with this second peak because if the ANMAT (that would be, here, your FDA), ask me what is it, I will not be able to answer...
I`ve read the USP method, but that involves a simple acid-base titration, useless with my matrix.

Tomorrow I will try to change the amount of reagent and I will do a new linearity at 60' - 80ºC (that`s when the second peak reaches it`s higher value), and see what happens...

Thanks for all

[HW Mueller]

I am a bit confused, with the cholic acid in the reagent mix you get a reagent peak + two more peaks? If so, it would be prudent to prove what´s in these two peaks before doing any more experimenting.

[Sentenced]

Well, that´s basically what I´m trying to solve here... The problem is that I don´t fully understand the behaviour of these peaks...
I have the results of the last 2 tests:

a) The first peak is linear at 80ºC - 60 min., but the second one isn´t.
b) If I double the concentration of the reagent, the first peak area decreases 2.5 times and the second one increases its area in the same rate.

So my idea now would be to try with a greater concentration of the reagent (4 times) and see if the first peak area reaches zero and I only have the second peak, and, if that´s the result, see if it is linear and stable along time.

Any suggestion?

[Mark Tracy]

If you assume that the molar absorbtivities of the DNPH and the derivative are nearly the same, you can estimate the peak area for the derivative from the molar concentration of DHC and the peak area and molar concentration of the reagent. You would need to repeat the calculation for 1:1, 1:2 and 1:3 stoichiometry. Does the area found come close to the estimates? This will tell you if you have complete reaction. I suspect not, but the numbers will tell.

[Bruce Hamilton]

My guess is that your aggressive conditions are transforming the Dehydrocholic acid between it's two polymorph forms ( alpha and the more stable beta ).

If that's the case, I'd suggest that it is a bad idea to change polymorphs during an analysis, as you may be invoking mechanisms that may change with sample. I'd again recommend minimising the acid concentration and the time/temperature.

Bruce Hamilton

[Sentenced]

I´ve verified the stoichiometry of the reaction and the reagent is 30 to 100 times in excess.

I´ve also tried increasing the acid concentration (before reading the B.H. message) and I´ve found that the first peak concentration is stable, but the second one increases with the concentration of DHC (at 80ºC . min), but it isn´t linear.

My question would be the same, if we have here a polymorphism, why, when the second peak area grows, the first peak area remains unchanged??

When I first started with this reaction, I´ve tried with 10 minutes at 40ºC, and I had poor reproducibility and low area values (about 5% of the area at 80ºC-60 min).

I don´t think I should consider the polymorphism, at least, based on this paper:

"Dehydrocholic acid in its a-form is a stable polymorph obtainable
by standard crystallization. (...) The b-form may be obtained, through a solvent-mediated transformation, by means of prolonged contact of the solid a-form with water–acetic acid (1:1) or water–HCl solutions
(at pH 3–4) under stirring. The time required for the conversion is either 3 or 40 days, depending on the solvent. On the other hand, the b-form may be reconverted into the a-homologue by heating above 228.5ºC (...). This thermal process is not reversible."

(Polymorphism of dehydrocholic acid: crystal structure of the
b-phase and guest-mediated solid phase conversion, Giancarlo Fantin et. al)

I´ll keep on trying... I´ll see what happens with more acid (I´ve tried with H3PO4, now I´ll change it for H2SO4), and with more reagent.

Lets see what happens!

[HW Mueller]

If I read your first post correctly you started the derivatization at a pH of below 1 which is lower than the pH used in your cited article. Bruce seems to be right?
Also, unless you do a mass balance you could run into trouble to interpret relative peak sizes in a problem like this. All of this makes sense only if you establish the identity of the peaks first, as I stated before (more bluntly: It seems that you are wasting your time with this reagent).

[Sentenced]

I´ve been reading that specific paper and it talks about the crystallization structure, but I have a solution, so it just doesn´t apply...

Of course, I'm having at least two products with this reaction, and it would be wonderful to have a MS, but I haven't... so that´s what I have so far... I would try something else, but I want to give it another chance to this technique, because the first peak is linear and I have 100% recovery with my matrix, I´m just trying to understand the behaviour of the second peak...

And, after all, this is the only thing that have worked, at least a little, because of the complexity of the matrix...

In about 2 hours I´ll have the results of the last tests... then I´ll see what happens with this...