Derivatization of a-hydroxy acids

[widetrak]

We are developing a method for determining lactic acid (LA) and salicylic acid (SA) in a mixture of collodium (cellulose nitrate) diethylether and ethanol. The method is based on this article http://journal.scconline.org/pdf/cc2002 ... p00126.pdf. The sample is dissolved in acid-DMF and derivatized vid BSTFA:TMCS. We use glycolic acid as internal standard.

Very straightforward and everything is working fine except for the recovery of LA in the product. The recovery is about 70% while SA is close about 100%. Spiked samples yield 100% recovery. Any suggestions on what the problem is?

We use EP-grade lactic acid in the product, is it possible that polylactic and lactates are not derivatized?

From EP:
DEFINITION
Mixture of 2-hydroxypropanoic acid, its condensation products, such as lactoyl-lactic acid and polylactic acids, and water. The equilibrium between lactic acid and polylactic acids depends on the concentration and temperature. It is usually the racemate ((RS)-lactic acid).
Content: 88.0 per cent m/m to 92.0 per cent m/m of C3H6O3.

[chromatographer1]

Your problem is the lactic acid you are using is not pure lactic acid, but as you have stated, it is a

Mixture of 2-hydroxypropanoic acid, its condensation products, such as lactoyl-lactic acid and polylactic acids, and water.

There will react with your reagent but they do not form TMS-lactic acid when they do. Thus, you are only finding the amount of lactic acid that is in your EP-grade lactic acid mixture.

This is not surprising. Salicylic acid does not have a sizeable amount of condensation products present and that is why your recovery is close to 100%.

You should run stds of the derivitized EP-grade lactic acid to determine the recovery of EP-lactic acid, Using high purity lactic acid for a comparison to EP-grade is not valid or 'fair' so to speak.

I hope I communicated this clearly.

best wishes,

Rod

[Don_Hilton]

The problem with lactic acid is that if you could get 100% pure lactic acid, the moment you have it, it goes to equilibrium of monomer, dimer, etc. plus water. The problem is catching the monomer and removing it from the equilibrium while maintaining the acidity to keep the reaction going. It crosses my mind that making a methyl or ethyl ester with a large excess of the alcohol might do the trick. And, if this is a valid approach, I would expect to see it in the literature - which I have not checked (as I am supposed to be fixing supper right now...).

Given that lactic acid is an important analyte, it might be interesting to see what is out there in the literature.

[KM-USA]

Yes, read the USP Monograph for lactic acid; a significant amount is present as in internal ester with itself. The USP assay saponifies the lactic acid sample with excess caustic, then back-titrates and calculates as lactic acid, essentially a traditional "saponification value" test.

We have a product with several percent lactic acid, and we saponify the sample and assay for lactic acid by HPLC using either UV or conductivity detector. I was aware of that paper, but decided to go this HPLC route for lactic. We use sodium lactate as the reference standard.

Edit: We have done salicylic acid by GC, but typically would use HPLC-UV.

[chromatographer1]

I think there is a point being overlooked. It is that

widetrak

appears to be doing a QA test, testing the finished product.

widetrak

wants to demonstrate that the product has been properly mixed and formulated. The test should demonstrate the amount of EP-grade lactic acid used and recovered in the finished product.

If a determination of pure lactic acid can be performed on the product, after converting the EP-grade lactic acid in the product, a similar test must be done on the original EP-grade Lactic acid used in the formulation.

The lactone (internal ester) and the di- and tri-mers, and their cyclic esters would all have to be converted to the free acid monomer of lactic acid. Not an easy job to accurately confirm and to validate in a method.

best wishes,

Rod

[KM-USA]

That's why I detailed our saponification-HPLC procedure. This has been GLP-validated.

[chromatographer1]

But it has not been validated in the product matrix we are discussing, a mixture of collodium (cellulose nitrate) diethylether and ethanol.

There lies the rub.

Rod

[widetrak]

Thank you for your input.
We have tried both acid and basic hydrolysis prior to the derivatization, in order to crack the esters. Basic hydrolysis works but it is not pretty, sometimes we get two phases and most often we get a precipitate.

I have also tried to make methyl esters both with methanol:BF3 and just plain caustic MeOH, extract with heptane and wash with sat. NaCl solution. The chromatography is poor however, it is impossible to identify the peaks. We also tried enzymatic test kits but ended up with the exact same results as for the GC-metod, that is 70% recovery of the lactic acid

Chromatographer1 suggests a simple route in his first response that I will try.
I would have considered HPLC but I think the sample work-up will be difficult considering the matrix.

[widetrak]

I have now tried to assay the product with the same lactic acid that is used in the formulation. Basically I used the method outlined in http://journal.scconline.org/pdf/cc2002 ... p00126.pdf but with the GA as internal standard and I also heated the prepared vials for about an hour. At first the results looked dissappointing but then I re-run the same vials in the sampler on the next day and now the results were very accurate. I have not had the time to see if I can reproduce these results yet, but form these results it seems that the reaction is very slow. How can I speed up the process?

[widetrak]

Yes, read the USP Monograph for lactic acid; a significant amount is present as in internal ester with itself. The USP assay saponifies the lactic acid sample with excess caustic, then back-titrates and calculates as lactic acid, essentially a traditional "saponification value" test.

.

We have a very simple method to determine the salicylic acid with a spectrofotometric method so I have been thinking about a method to titrate "total acid" so I will take a closer look how to determine a SAP-value. Thanks for the advice.

[magda roth]

I have now tried to assay the product with the same lactic acid that is used in the formulation. Basically I used the method outlined in http://journal.scconline.org/pdf/cc2002 ... p00126.pdf but with the GA as internal standard and I also heated the prepared vials for about an hour. At first the results looked dissappointing but then I re-run the same vials in the sampler on the next day and now the results were very accurate. I have not had the time to see if I can reproduce these results yet, but form these results it seems that the reaction is very slow. How can I speed up the process?

I cannot get the article you are quoting, is there any way you could e-mail it to me? Somehow the link does not work (or I am again proven not to be good at computer stuff )
I am also fighting lactic acid in a rather complex matrix.
Thanks
magda.roth@gea.com

[dblux_]

I cannot get the article you are quoting, is there any way you could e-mail it to me? Somehow the link does not work (or I am again proven not to be good at computer stuff )

It has nothing to do with your computer skills.
After almost a year this link is not valid.

Below active link:

http://pubget.com/paper/11919646/Simult ... matography

[KM-USA]

I just clicked on the link in the very first post, it worked for me.

Consider sodium lactate as the standard for lactic acid.

[dblux_]

I just clicked on the link in the very first post, it worked for me.

...

Yes, server is on again.