Extraction method for doxycycline

[tlili]

Hi all,

I am trying to determine the % recovery of doxcycline in human serum. When I run a blank (pure plasma), the dirty peaks eluted at a RT of 2.4 min to 2.7 min that was very near to my analyte peak (RT 2.7 min). At higher concentration of doxcycline, I am still able to get my analyte peak. But at low concn (2 ug/ml and below), I am unable to get the analyte peak.

I am using liquid-liquid extraction method. The organic solvent that I was using in the whole extraction method was ACN

Mobile phase composition: 20ME 55 0.15% phosphoric acid 25ACN. Isocratic. Injection vol: 30 ul at 1 ml/min
Column: Symmetry shield RP18 4.6 x 250 mm
wavelength: 347 nm

Appreciate if anyone can suggest how I can overcome this problem? By the way, how do we determine what sort of organic solvent to be used in extraction?

[bert]

In fact, your extraction with acetonitrile is a protein precipitation step. Did you try to improve sample clean-up, using solid phase extraction?

Regards Bert

[Uwe Neue]

I agree with the comment by Bert. Your method is not very good for getting a clean sample. However, and much more important at the moment, your analyte peak is nearly unretained, and this is the reason for the strong interference from the serum constituents. Lower the acetonitrile concentration!

I am not sure if I understand the composition of your mobile phase. Assuming that I read it correctly as containing 25% acetonitrile, go to at least 20% acetonitrile, and maybe to 15% acetonitrile. The retention of your peak will increase significantly, and maybe this will be good enough for separating it from the interferences.

If this is not good enough, I recommend to use solid-phase extraction. There are some really good and well proven methods around. You can look at the Waters web site under Oasis, or I can send you a review article with lots of background information.

[tlili]

Bert and Uwe Neue, thanks for all your suggestions.

Uwe Neue,

Really appreciate if you can send me those review articles w lot of background information.

But just some clarifications, if I were to lower the % of ACN, do I increase the % of MeOH or 0.15% phosphoric acid? Which one is better? Is there any principle to base on?


Bert,

What is the purpose of protein precipitation step?

[bert]

You cant inject pure serum on your LC system. Due to the proteins it would clog. That is why you deprotenate the samples before injection.

If you need more retention you could also use heptane sulphonic acid (HSA) as an ion-pair reagent. In that case you need a low pH of your mobile phase (try 2 - 2.5) and you probably need 1 to 3 grams HSA per litre mobile phase.

[HW Mueller]

There exist restricted access columns which are designed for injecting plasma/serum directly. If you are lucky this may work for you (I have used them successfully only as a first step in multi-step HPLC).
If you have to stick with the solvent extraction (really a solid-liquid extraction) you have to use one that precipitates the proteins (to remove them, as mentioned), but it also must dissolve your unknown (don´t forget that the water is also extracted). If it is not dissolved it will just stick to the protein or to the vessel´s surface (if you use trichloroacetic acid this is very likely to occur). Also one has to shake the solid/liquid (protein/H2O-organic) vigorously, for a long time, to prevent the coprecipitation.
Another possibility is a ultrafiltration to remove the proteins, but one usually has to add a small amount of organic to keep ones unknown in solution, rather than attached to the proteins. The lower amount of organic solvent used here tends to result in a cleaner chromatogram as compared to the liquid extraction.
A carefully executed two step chromatography (as Uwe suggested, but restricted access should also be considered as first step) could give a final chromatogram with the least interference.