Residual ethanol in tablets

[Mattias]

Hi,

We are analysing residual ethanol in tablets with headspace GC with FID detection.

We have problems to get the ethanol out of the tablets in a reproducible way. We dissolve the tablet fragments in water.

Anyone experienced the same behaviour? My theory is that the ethanol is bound to the starch. Any suggestion for improvement is appreciated!

[Peter Apps]

What are the symptoms of poor reproducibility of extraction, and how are you discriminating between this and, say, poor repeatability of the partition of ethanol into the vial headspace ?

What headspace system do you have ?

Peter

[Mattias]

This analysis is carried out in a contract lab, but I would like to improve the method.

There is a huge difference in RSD when analysing three samples - compared to three standards. That is why I suspect the sample preparation to be unrobust.

The equipment they are using is from Agilent - but I am not sure of the model numbers right now.

[chromatographer1]

Ethanol does not bind to starch in water.

If the drug containing ethanol being capsulized is not homogeneous then the recovery of the ethanol from the tablets will also not be homogeneous.

This, for example, often happens when drug material is dried in pans. The corner material will be 'wetter' than the middle.

Bad (inconsistent) results from the tablets do not necessarily mean your analytical method is poor. It may mean that your samples have poor consistency. Which is the purpose of the analytical test, is it not?

best wishes,

Rod

[Peter Apps]

Rod is right - you are not measuring repeatability if you analyse three different tablets becuase you do not know that all the tablets have the same ethanol content.

If you have good repeatability from standards (I presume that this is ethanol in water ?) and different samples give you different results, the test that shows whether this is due to unrepeatable extraction or to real differences between samples is to grind up a batch of tablets together, carefully homogenise the powder and then analyse portions of the powder.

Peter

[Mattias]

You may be right. I am hestitant to grind the tablets since I fear that the ethanol will evaporate - but I will try to get a more homogeneous sample just to test the method.

The granulate is dried in a "one-step granulator/drier". It is also mixed with Mg-stearate before tabletting. It should quite homogeneous, but nothing is certain!

[chromatographer1]

Grinding is not necessary.

Prepare a solution large enough to dissolve 5 or 10 tablets without a headspace in the container.

Then quickly sample this solution and do multiple tests of the solution to determine the 'real' amount of ethanol averaged over 5 or 10 tablets.

Your standard addition of ethanol to the sampled solution will also give you an idea of the linearity and repeatability of your test.

You must determine if your test is reliable first. This can be done with heated solutions that have had all ethanol removed and then spiked with known amounts of ethanol for testing of reproducibility and linearity.

In other words, use the proper dissolved tablet matrix to determine these factors.

best wishes,

Rod

[Tomasz]

As I remember the limit for ethanol is very high and thus the level of ethanol in tablets is usually much lower. Do you use ethanol standard at concentration level similar to this found in sample solution or higher? If you have more ethanol in standard than in sample solution, the RSD of standards could be lower than RSD of samples.

Regards