Mystery peak showing up in Atropine Assay

[KarenJ]

I recently began running the USP method for Atropine Sulfate injectables, which involves an isocratic run of acetate buffer with tetra-n-butyl ammonium hydrogensulfate with a pH=5.5 mixed with Acetonitrile (20:1). The Atropine Sulfate injectables have 2mg of Atropine sulfate formulated with glycerine, phenol and a citrate buffer, the samples are prepared by adding the samples into a 25-mL Vol. flask and diluting with water. To date the first few injections work adequately, but when I get to injection #4 I begin to have a mystery peak with a relative retention time of 0.47, and relative areas around 90%. This peak will then continue to show up in all subsequent injections. However, if I add a blank after each sample run the peak does not appear in the blank, but will appear in the next injection of the sample. I have also added an injection loop wash, which is done before and after each injection, and the peak continues to appear starting with the 4th injection of the sample. I have also found that rearranging the order of the samples, e.g. sample #7 is injected first while sample #1 is injected last, still results in the peak appearing in the 4th injection. This peak has not been found to be present in any of the USP std runs that I have done. Does anyone have any ideas of what or where this peak is coming from, or ideas of how to get rid of it? Thank you.

[danko]

My guess would be: The peak is “realâ€

[AdrianF]

I would try making up my sample in mobile phase, this is always a good idea when you have an ion pair mobile phase.

Another possibility is to use a more dilute sample - the main peak can have an absorbance of 0.01AU with no loss of acuracy. Those of us who have worked in trace analysis have had to get sensible results with a range of 0.001AU.

Try other columns.

Danko gives a good method of diagnosis but I don't suppose you can do anything about the formulation.

[danko]

I don't suppose you can do anything about the formulation.

Maybe not. But is it not the role of the laboratory – finding/identifying potential problems?
It’s certainly not hiding them. Otherwise I wouldn’t be particularly comfortable with using this drug.
Besides, I can’t see how dissolving the drug in mobile phase would help. And frankly, I don’t hope it will.

Best Regards

[AdrianF]

I notice the USP method involves a 100ul injection, you could try a 10ul injection.

If you are not bound by the USP method you could try the BP method below.

If you must use the same method, investigate different columns, the L1 designation in the USP could be any one of 220 or more columns. Which column are use using now?

The chromatographic procedure may be carried out using (a) a stainless steel column (10 cm × 4.6 mm) packed with end-capped octadecylsilyl silica gel for chromatography (5 µm) (Nucleosil C18 is suitable), (b) as the mobile phase with a flow rate of 2 ml per minute a solution containing 0.01M sodium acetate and 0.005M dioctyl sodium sulphosuccinate in methanol (60%) adjusted to pH 5.5 with glacial acetic acid and (c) a detection wavelength of 257 nm.

PS. I realise the tetra-n-butyl ammonium hydrogensulfate is not acting as an ion-pair reagent in this assay however a 100ul injection of water may disturb the equilibrium of this this reagent which is being used to supress the effect of silanols.